Retail Payment Systems
In the OIC Member Countries
18
one of these externalities continues to exist, the market is two-sided, though the intensity of
the externalities may vary across payment systems, countries and even within countries.6
Adoption externalities relate to the fact that a payment platform with more payment cards
issued is relatively more attractive to merchants than one with fewer cards issued. Similarly, a
payment system with many merchants registered may be more attractive to consumers than
one with fewer merchants. Reaching a minimum scale for success may require years of
investment including much sunk cost. Moreover, there is no guarantee of success, even by
payment systems that would be successful if they reached sufficient scale.
Unarguably, retail payment systems represent a substantial part of the whole economy. For
instance, as early as 2000, Humphrey et al. (2000) estimated that the United States spent $225
billion per year just to make payments, or 3% of GDP.7 For other countries, the percentage of
GDP is likely to be lower: these countries generally use cheaper instruments than cheques and
the number of non-cash transactions is lower than in the US. Meanwhile, De Grauwe et al.
(2000) do include the cost to the merchant and get a cost of 0.7% of GDP for cash alone. Some
0.2% is borne by banks, leaving 0.5% to be borne by consumers and merchants. Overall these
estimates indicate that the cost of the payment system represents something in the range of
1.5-3% of GDP for developed countries. It is also expected that the cost of the payment systems
in developing countries will be slightly larger. In most cases, this is lower than the cost of
operating cash systems, where the burden is widely distributed as risk, and associated
insurance and security expenditure, to all users and especially small businesses.
Indeed, costs vary significantly across the various instruments. Unfortunately, most of these
costs are hard to measure: they represent processing costs for banks, and handling costs of
merchants, businesses and consumers—none of these are readily obtainable.8 The impact of
using different instruments can also be large. For example, Humphrey et al. (2000) estimated
that the United States could save 1.25% of GDP in the move from paper cheques to electronic
giro. Similarly, Humphrey et al. (2001) estimate that Norway could save 0.6% of GDP by
moving all of its paper-based instruments to electronic versions. Unfortunately, to the best of
6 The price structure will develop in such a way that both sides have incentives to participate in the market. Rochet and
Tirole (2003) call this a two-sided market. They state that the volume of transaction on and the profit of a platform depend
not only on the total price charged to the parties to the transaction, but also on its decomposition.
7 This huge figure is based on an average cost per non-cash transaction of $2.60 times 87 billion non-cash transactions per
year. It excludes the cost of cash, but it still uses the old check estimate by whom? (66 billion cheques per year, which was
later revised downward by 17 billion.
8 Several authors have tried to estimate costs of the main payment instruments. For example, Humphrey et al. (2003) and
van Hove (2004) give a comprehensive overview of the state of the art in estimating payment costs. Both highlight the
pitfalls and difficulties in obtaining estimates that are comparable across countries.