Is money demand really unstable? Evidence from Divisia monetary aggregates

We revisit the issue of stable demand for money, using quarterly data for the European Monetary Union, India, Israel, Poland, the UK, and the US. We use a modern version of the same linear time-series macroeconometric modeling and specification approach that had previously cast doubt on money demand stability. Autoregressive distributed lag (ARDL) cointegration models are used in the study to establish a long-term relationship among real money balances, real output, interest rate, and real effective exchange rate. For all the countries analyzed, evidence of stable demand for money is found. Broad money in general is better at capturing a stable demand for money than narrow money. The stability results are especially strong, when broad Divisia money is used instead of its simple sum counterpart. Our results are consistent with the large literature on the Barnett critique, which is based on a different methodological tradition that employs microeconometric modeling of integrable consumer demand systems. That literature has never found the demand for monetary services, measured using reputable index number and aggregation theory, to be any more difficult to model or less stable than the demand for any other good or service in the economy.