Financial Exclusion and the Distributional Limits of Monetary Policy

In economies where a portion of the population transacts through mobile money and the other portion strictly uses only cash, can any single interest rate rule serve both groups well? I develop a two-agent New Keynesian model calibrated to Ghana in which included households manage liquidity through mobile money under Baumol– Tobin demand, while excluded households depend on government transfers under fiscal dominance. I find a critical threshold at approximately 70 percent financial exclusion.Below it, aggressive inflation targeting is optimal for both household types. Above it, the welfare surface for included households develops an interior optimum, the optimal Taylor rule diverges across groups, and no single rule resolves the conflict. The distributional cost of monetary policy is convex in exclusion: the welfare variance ratio between household types rises from 7.6:1 at 50 percent exclusion to 98:1 at 80 per- cent, the range observed across Sub-Saharan Africa. Aggregate welfare statistics mask this entirely. The trade-off is reducible only through financial inclusion, not through monetary policy design.