|
Article Excerpt
"Except for the smallest of transactions, money will no longer be a physical thing." (Forbes 1967 as cited in Flannery 1996)
"The use of cash and currency will drop drastically ..." (Flannery and Jaffee 1973)
"Cash is dirty, inefficient--and obsolete." (Gleick 1996)
"The end of the cash era." (The Economist 2007, cover page)
WHILE MANY OBSERVERS have predicted the replacement of cash for decades, cash holdings remain significant and show few signs of decline in the aggregate. During the last 15 years, retail electronic payments have grown rapidly in most advanced countries. More recently, many traditionally cash-only establishments such as fast food restaurants, vending machine operators, and taxis have started to accept cash alternatives resulting in greater downward pressure for transactional cash demand. If there are more merchants accepting cash alternatives and more consumers using them, why is the stock of cash still so high? A significant part of the answer is that cash is not solely used as a means of payment.
In this article, we are primarily interested in answering the following question. Given greater usage of cash alternatives, are households and businesses holding less cash to make and accept payments! We disentangle cash's transactional role from its store of wealth role by categorizing currency into different denomination classes--small, medium, and large. We find evidence that greater usage of debit cards indeed reduces the transactional demand for cash--most notably, the demand for smaller denomination currency notes and coins. In addition, we find that greater consolidation in the retail industry also contributes to lower transactional demand for cash, suggesting that larger merchants may be better able to absorb the fixed costs of accepting electronic payments.
This question is of interest to economists and policymakers for several reasons. Most directly, greater usage of cash substitutes affects how much cash the central bank should supply and also influences the income and interest rate elasticities of money demand. The implications of a possible reduction in currency demand for monetary policy has been an active area of economic research (Friedman 1999, Freedman 2000, Goodhart 2000, Woodford 2000, Kroszner 2003, Chakravorti 2006, Alvarez and Lippi 2007). Moreover, some economists have suggested that social welfare would improve if fewer cash transactions took place (De Grauwe, Buyst, and Rinaldi 2000, Van Hove 2004). Finally, there has been considerable debate among economists regarding the role of financial incentives on the adoption of electronic payments and its impact on the demand for paper-based payment instruments (Humphrey, Pulley, and Vesala 1996, Chakravorti and McHugh 2002, Markose and Loke 2002).
We discuss payment trends in 13 Organisation for Economic Co-operation and Development (OECD) countries--Austria, Belgium, Canada, Finland, France, Germany, Italy, Japan, the Netherlands, Sweden, Switzerland, the United Kingdom, and the United States--from 1988 to 2003. We find that in most countries, debit card usage has increased significantly. However, the aggregate demand for cash has not fallen drastically with widespread adoption of noncash payment instruments. Unlike other payment instruments, cash serves as a universally accessible store of wealth in addition to being a payment medium. Thus, any analysis of the impact of cash alternatives must first disentangle the two competing functions of cash. Furthermore, in contrast to payment cards, cash transactions are more difficult to trace to the transactors because it can be used multiple times without third-party intervention. Consequently, cash will continue to be attractive for facilitating illicit transactions, as well as for storing ill-gotten (or, at least, untaxed) wealth.
Our analysis differs from the existing literature in a number of ways. First, we decompose cash demand into three currency denomination categories to better identify the transactional motive for cash holdings. Denominations larger than those commonly dispensed by a country's automated teller machines (ATMs) are posited to be better suited for store of value purposes. Denominations less than those dispensed from ATMs are required to make change for purchasers that obtained currency from ATMs. We argue that since change is given only in the case of cash transactions, the effects of substitution toward electronic payment instruments are reflected primarily in the demand for these small-denomination notes and coins. If cash is not used to transact, the demand for change and consequently the demand for small-denomination currency both decrease. Second, we employ different econometric techniques from those commonly used in this literature to address strong serial correlation that is present in annual cross-country payments data. Finally, our data cover both the time of rapid growth and subsequent saturation in debit card usage in almost all countries in the sample. That is, unlike many of the previous studies, we are able to analyze the effect of accelerating and decelerating changes in debit card usage on cash demand.
We find that the effects of increased usage of debit cards are indeed confined to demand for small-denomination currency and coins. Both debit card terminal proliferation and financial institution branch infrastructure significantly affect the demand for small-denomination currency, whereas the short-term interest rate (STIR) has little effect. Because financial institutions are the main distributors of small- (and large-) denomination notes and coins, more extensive branch networks are associated with greater demand for small- (and large-) denomination currency.
In contrast, the demand for larger denomination notes is affected by the STIR, but not by the adoption of point-of-sale (POS) debit card terminals. Moreover, there is some evidence that greater adoption of ATMs has, on net, reduced the demand for ATM-dispensed notes. We also find that the share of small merchants in the economy (as proxied by the share of the self-employed) is associated with stronger currency demand. However, this effect is primarily observed through demand for small-denomination currency, suggesting that high fixed costs of installing electronic payments terminals and not illicit activity influence cash demand by small merchants.
This paper is organized as follows. In Section 1, we discuss payment trends across the countries in our sample. In Section 2, we review the literature that analyzes the impact of electronic payment alternatives on cash. In Section 3, we focus on the relationship between debit card usage and changes in the stock of cash, broken by denomination categories. We offer some concluding remarks in Section 4.
1. CROSS-COUNTRY PAYMENT TREND COMPARISONS
While there are major differences in banking and regulatory structures, as well as in cultural attitudes toward payments between countries in our sample, certain common payment trends exist. Figure 1 depicts one such common trend--the growth in debit card usage during the 1990s. In 1990, all countries except for Finland and France had fewer than 10 debit card transactions per person per year. By 2003, the average number of transactions approached 50, with 5 out of 13 countries recording more than 70 debit card transactions per capita. This monotonic trend suggests that the net benefits of using debit cards have increased vis-a-vis other payment instruments for both consumers and merchants (assuming that the total number of transactions is growing at a slower rate than transactions executed with debit cards).
[FIGURE 1 OMITTED]
Another trend is the rapid growth of electronic payments generally. As shown in Table 1, even in countries with a historically high share of check transactions (e.g., the United States, France, and Canada), electronic payments accounted for a vast majority of all noncash payments in 2005. (1) Furthermore, as also reported in Table 1, noncash transactions have increased between 2001 and 2005 for all countries in our sample, suggesting a reduction in cash transactions over time within a country.
However, the growing acceptance of cash substitutes has not necessarily resulted in an offsetting reduction in the total stock of currency outstanding. Two measures to compare currency in circulation across countries are cash holdings per capita and the ratio of cash outstanding to GDP. Both of these metrics show substantial cross-country heterogeneity that does not disappear over time. Looking at the opposite ends of the spectrum, the 1988 per capita currency holdings in the United Kingdom amounted to $529, while those in Japan were $2,296. The corresponding figures in 2003 (measured in 1988 prices) were $671 and $5,292. The currency-to-GDP ratio in the United Kingdom changed little over this period, going from 3.5% in 1998 to 3.4% in 2003. In contrast, this ratio for Japan rose from 9.2% to 16.3%...
|
