The Impact of Technological Change on the Currency Behavior of Households: An Empirical Cross-Section Study

Article excerpt

In recent years, there has been a substantial change in the arrangements by which households can manage their transactions balances. The removal of Regulation Q and the introduction of NOW accounts has changed the attractiveness of holding transactions balances and influenced the relationships between the monetary aggregates and other macroeconomic variables. There has been a simultaneous change in the deployment by banks of over 60,000 automatic teller machines (ATMs) and the acceptance of direct deposit of payroll and other payments to households, especially Social Security payments. These technological developments presumably affect various transactions costs of households in managing their transactions balances. In particular, these developments have important implications for the demand for currency by households. It has been shown that sophisticated corporate cash balance management techniques, making intensive use of technology, have affected the demand for money (Kimball 1980 and Dotsey 1984). There has been little systematic study of the impact on household behavior, however. The purpose of this paper is to analyze the impact of technology on the currency inventory behavior of two large samples of households in the United States. If technology affects transactions costs of obtaining an inventory of currency, then inventory theory would suggest that this should lead to more transactions and lower average inventories of currency, ceteris paribus. This hypothesis is tested in this paper.

In section 1, the model is developed. In section 2, the data source and the empirical findings are discussed. Section 3 the subject is summarized and concluded. Using data from two large surveys in the mid-1980s, the results of this paper indicate that technology has a substantial impact on the behavior of households in holding currency inventories.


The theory of the demand for currency has its roots in the theory of the demand for money first put forth by Baumol (1952) and Tobin (1956). These contributions adapted developments in inventory theory for examining the demand for money. In general, the demand for money was viewed as related to the amount of expenditure to be financed (the scale variable), the cost of a transaction between money and an alternative interest-bearing asset ("bonds"), and the interest return on the alternative asset. The money inventory was directly related to the scale variable, but there were economies of scale, the elasticity of demand being exactly 1/2 in a deterministic model. Higher interest rates raised the opportunity cost of holding money, leading to lower inventories, while higher transactions costs of exchanging money and "bonds" would lead to higher inventories. An explicit model of household money demand was developed by Santomero (1974) in which the traditional definition of "money" was expanded to include both bank deposits and currency. This type model was recently applied to aggregate data by Dotsey (1986, 1988). Most of the empirical studies of the past twenty-five years have utilized time-series aggregate data (Cagan 1958; Hess 1971; Fry 1974; Becker 1975; Lewis and Breen 1975; Garcia and Pak 1976; Boughton and Wicker 1979; and Ochs and Rush 1983). Of course, while aggregate time-series studies are useful, cross-section studies provide a more direct test because they utilize individuals and households as units of observation and thereby avoid the serious problem of aggregation bias.

In this paper, the household is viewed as making a hierarchy of decisions. First is the choice between consumption and saving, and we take this choice as given. Next, the household chooses between different payment media in making purchases. The choice among paying by check, currency, or with a credit card is determined by relative costs and benefits, including such considerations as the time spent in conducting the transaction, the risk involved in carrying the different payment media, the need for a payment record, and the desirability of deferred payment. …