Working Paper: CEPR ID: DP2818
Authors: Olympia Bover; Nadine Watson
Abstract: We estimate scale elasticities in firms' money demand using panel data. Our main data set is a sample of Spanish companies observed over 1983-96. We also analyse comparable UK and US data sets.We find that the errors in money demand equations contain two terms correlated with sales: first, a permanent firm effect capturing differences in managerial efficiency, efficiency wages, technological sophistication; second, a measurement error in sales, probably because cash holdings are end-of-period whereas sales are annual measures. We show that failure to control for them results in important biases.Sales elasticity estimates for Spain increase substantially jointly considering correlated fixed effects and measurement error. Additionally, our estimates indicate declining sales elasticity from mid-1980s to mid-1990s, a period of increasing financial innovations. This suggests that financial innovations reduce money demand mainly by reducing the sales elasticity.We also estimate interest rate elasticities using both aggregate and firm specific rates.
Keywords: firms; money demand; measurement error; panel data; technological change
JEL Codes: C23; D21; E41
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Sales (L81) | Money Demand (E41) |
| Correlated Fixed Effects and Measurement Error (C33) | Sales Elasticity (D12) |
| Financial Innovations (G29) | Sales Elasticity (D12) |
| Interest Rates (E43) | Money Demand (E41) |
| Financial Sophistication (G53) | Interest Rate Sensitivity (E43) |