Why is the Use of Cash Persisting?

Cash has been the most long lasting success story in payment history. Paper money was recognised as legal tender in Germany over 100 years ago and today over 75% of retail payments are still made using cash.1,2 In Europe, cash payments accounted for 78% of the continent’s 388 billion retail transactions in 2008.3 In Eastern Europe, cash represents 93% of payment volumes and is thus almost a monopoly. While in 1913 each person statistically carried only one banknote, today on average 37 notes are carried per person, far more than the number of credit, debit or other cards carried. The volume of cashless payments in Europe grew by 62% (CAGR 6.2%) between 2000 and 20084 or even faster when considering normal payment bills in circulation (see Figure 1).

Figure 1: Cash in Circulation Increasing Globally (Only Non-hoarding Banknotes Shown) Despite Increasing Alternatives

Source: CapGemini World Payments Report (2010)


All this shows the unprecedented ‘success’ of cash (in the biological sense of proliferation, avoiding predators and spreading of its genes) in the worldwide payment ecosystem.

The financial inefficiencies of handling cash are also well known, however, and have been analysed in even more depth only recently.5 The total cost of distributing, managing, handling, processing and recycling cash and of accepting cash payments within the borders of the euro system was €84bn in 2008 – the equivalent of 0.6% of Europe’s gross domestic product (GDP) or €130 per person.6 Worldwide cash handling costs total more than $300bn per year.7 To put this into perspective, the illicit drug trade is estimated by the UN at US$320bn a year.8 Indeed, there are more parallels between cash and the drug trade, beyond representing similar sized monetary volumes.

Cash is the basis for the underground economy and a solid majority of US bills are contaminated by cocaine (directly from drug users sniffing, indirectly by being spread through ATMs). It is also unhygienic: 94% of dollar bills harbour staphylococcus (thus carrying more germs than a typical household toilet).9 This is due to the fact that a paper bill being carried in a warm, possible moist, pocket is a hospitable environment for viruses and bacteria (which can live on most surfaces for about 48 hours, but live on paper money for up to 17 days). Thus, cash is not only primitive and costly, it is also unhealthy.

Focusing only on the cost aspect, it has been shown10 that most of the cost occurs at the front end, where the single highest cost driver is personnel, accounting for 72% of costs in retail stores and 61% in banks. It is thus to front-end alternatives that one must look in order to reduce the costs of cash significantly.

The banking industry, whose share of the cash handling burden amounts to €25.6bn per year11 has undertaken many initiatives to reduce this front-line cost. Online banking and payment cards offer more efficient ways of making payments. But optimising cash logistics (e.g. ATM cash pay-in, geographical optimisation of banknote distribution), while making the handling of cash more efficient is actually an example of the German saying ‘Zementierung von Eselspfaden’ (casting mud tracks in concrete), which serves to perpetuate an inefficient instrument even longer – much like sending cheques by electronic image in the US.12

Not only banks but also many regulatory interventions (in Europe on subjects such as the single euro payments area (SEPA), Payment Services Directive (PSD), e-Money directives and electronic invoicing (e-invoicing)) are expected to encourage a move towards electronic payments (e-payments). The fact remains, however, that the value of cash in circulation in the eurozone is increasing, not diminishing (see Figure 2). This ‘healthy’ growth rate of 10% per annum (p.a.) means that the euro cash in circulation has almost doubled since its introduction in 2002.

Figure 2: Cash in Eurozone Growing at 7-10% a Year

Source: ECB


In some areas, however, the ‘war on cash’ does seem to be successful: debit card spending recently overtook cash transactions for the first time in the UK.13 Spending on debit cards reached £272bn in the 12 months to October 2010, while cash transactions stood at ‘only’ £269bn.

Cash machine withdrawals also dropped during the period, as consumers used their cards (also for smaller value transactions) at point-of-sale (POS) in shops. Contrast this, however, with Germany, Italy and Spain, where debit cards are used primarily at ATMs – to withdraw cash.

This is an indication of a wider point that cash usage – like that of all payment instruments – differs widely by country.

Figure 3 shows that, while European economies tend to have a lower share of cash transactions (shown here per volume) than many in the developing world, several countries in Europe tend be ranked at best ‘average’ (Italy, Germany) in a field where even the champions (Nordics, France) have only managed to eliminate half their paper-based money.

Figure 3: Cash Remains Important Everywhere

Source: McKinsey Global Payments Map


The US appears to be a cash-poor champion until one remembers that cheques are still incredibly prevalent (particularly, surprisingly, in business-to-business (B2B) transactions), a paper-based payment method no better than cash. Analysts14 estimate that 10.6 billion B2B US cheque payments were made in 2009 compared with only 1.99 billion B2B automated clearing house (ACH) transactions. Even given the relatively low level of cash usage in the US (where cash usage is already comparably lower and actually falling), even if the use of cash were to continue to decline by 17% every five years (the current forecast15) the use of cash would not fall below US$1bn before the year 2205, roughly 200 years from now.

Broadly summarising the findings in this section, it has been established that, despite many valiant efforts of regulators and banks and despite winning the occasional tactical victory, cash will remain the predominant payment method for many years to come.

Indeed, the public seem to have a Mephistophelean pact with the devil that is cash. As Goethe made Faust say about banknotes in 1832: ‘’Twere hopeless now the flying leaves to stop/With lightning speed they spread through-out the land’.16 Goethe, besides being a poet, was finance minister at the Court of Weimar, so he was well placed to comment on these developments, and his insights remain astonishingly relevant today.

Analysis of the Status Quo

In order to understand the resilience of what appears to be such a primitive, slow, expensive and potentially even hazardous payment instrument, it is necessary to examine the interests of the key stakeholder groups invested in the cash business. These are primarily the retail banks, central banks, national governments, official businesses such as supermarkets, unofficial businesses in the shadow economy and consumers.


Looking at the stakeholders from the financial services sector, one can identify two main stakeholders: central banks and ‘normal’ banks servicing customers and merchants.

Central bank studies (performed since 2001 in Belgium, the Netherlands, Norway and Portugal) all concluded with the surprising result that cashless payments have a significantly higher unit cost than cash payments. This is due to the lower volumes and the high infrastructural and operational costs of e-payments (e.g. POS terminals, networks, computer systems and fraud management).

The central bank studies all agreed, however, that the substitution of cash and paper-based payments by electronic account-based methods would reduce the societal cost of payments. But some of these central bank studies (Belgium, Netherlands) showed that the gross societal cost savings by realistic cash substitution were marginal at just 0.02% of GDP.17

Another side of the ‘coin’ is that central banks benefit hugely from the large amount of cash in circulation. The euro system currently has 14,200,000,000 bank notes outstanding, a total value of €447bn.18 All these notes and coins in circulation are not earning interest, thus saving the banking system tens of billions of euros p.a. (e.g. 3% p.a. of €447bn equals €13bn p.a. in savings).

It is thus apparent that central banks have a dual, heavily conflicting role: on the one hand, they have a mandate to foster the efficiency of payment systems; on the other hand, they are market participants as issuers (and heavy economic beneficiaries) of cash – the most inefficient payment system of all.19

Interestingly, the large denomination notes (€500, €200 – never seen in wallets) make up 40% of the value of cash in ‘circulation’. The large denomination bills in the eurozone far exceed the value available in other geographies (US$100/€73; UK£50/58; JPY10,000/€82; 5000 roubles/€120 as largest bills – all much less than €500). Since Europe has such significantly higher compact means of transporting money also in the shadow economy, the euro is in danger of becoming the currency of choice in the criminal underworld.20 This has led to the suggestion that the most effective way to reduce money laundering – instead of current bureaucratic, administrative and technical measures – would be simply to abolish the €500 and €200 bills. For this very reason, Canada withdrew its C$1,000 note from circulation.

One innovative (if slightly eccentric) idea to allow the continued use of banknotes by law-abiding citizens while making life difficult for the shadow economy is to implant radio tags into banknotes.21 This allows bills’ storage and movement to be traced and makes it impossible for kidnappers to demand ‘unmarked’ bills. Unlike serial numbers or bar-codes these RFID tags can be read for a whole bundle at once and at a distance. The tracking and tracing of the bills clearly raises severe privacy issues – but that is the point: ‘no more carrying suitcases full of cash to Switzerland’. The technology allows individual bills to be cancelled remotely, which is useful against ransom money but removes the fundamental property of ‘fungibility’ of cash. The proponents of the idea suggest, tongue-in-cheek one would hope, that this technology is also ‘great for instantly taxing citizens’, as the government can take the correct sum straight out of each taxpayer’s pocket by cancelling banknotes. This may be a step against tax evasion that has gone too far however. But considering that of £250bn estimated to have been laundered through the UK, of which the government has recovered a mere £46m (i.e. 0.02%), at a cost of £400m,22 we can see that drastic measures are necessary to combat the fraud inherent in cash and for which we all pay indirectly.

Many other means of curtailing the negative effects of banknotes have been suggested: restricting the number of ATMs, charging for cash withdrawals and replacing notes by high value (and heavy) coins. Explaining cost-based pricing to consumers is not easy however, and making cash more expensive is not going to win elections.

Turning now to ‘normal’ banks that deal with consumers and merchants, they suffer a high cost of cash. They are able to recuperate some of this from merchants (charging them cash withdrawal/deposit fees, coin issuance/counting services and all-night money deposit services). Also some banks make a good income with ATM cash services, particularly from ATMs of banks that are not their own. But since cash is free by law to consumers and since European law is eroding the prices for ATM cash services, a large and growing burden of cash handling remains with the banks. In order to compensate for this, banks are increasingly going some way towards optimising their cash handling at both the back end (cash centres) and the front end (ATM cash recycling). Beyond the financial burden of cash handling, one must also recognise the non-monetary interests that banks have in cash, such as the customer relationship, which is often based upon – particularly elderly and wealthier – customers regularly getting their money personally at the branch.

Finally, both central and retail banks employ an extensive service industry as suppliers in the money industry (be it coins or bills) who also have a vested interest in the longevity of cash.

Shadow economy

The large denomination bills identified above are of particular advantage to the European underground economy, as a million dollars weigh 10 kg, whereas one million dollars in euros weigh about 1.6 kg.

This is due to the fact that Europe has such large denomination bills, making cash transport much more compact (author’s own calculation). The mafia, for example, therefore prefer these large bills, because they can transport larger amounts of cash in less space.23

Cash payments which – unlike card, cheques or account-based transfers – are anonymous represent the basis for about 90% of all transactions in Italy, about 55% in France and about 75% in Germany. Italy has one of the largest underground economies in Europe, worth as much as 19% of GDP in 2008 (€300bn) – showing the correlation between high cash usage and high activity of underground economy.24 Even Norway’s cash is largely (71%) used for illegal purposes.25

The 566 million €500 notes in circulation outnumber the total population of the eurozone. There is a greater concentration of €500 notes per capita close to the borders of Switzerland and San Marino, where money-laundering regulations are less stringent.26

Thus the underground economy is a major stakeholder in cash and a major beneficiary from the high denomination euro notes.

National governments

Closely linked to central banks are national governments and their treasuries. These benefit from cash through seigniorage. The extent of the advantage is surprising, as illustrated by the following single example of many. The US treasury estimates that it has earned about US$4.6bn in seigniorage from issuing one coin collection alone, the ‘50 State’ series of quarters. Each quarter costs the mint less than 10 cents to make for the 147 million people who collect these sets of 50 quarters (at a price of 25 cents per coin).

Of course, cash causes great expense, huge loss of tax revenues and immense societal costs to governments through the consequences of the activities of the shadow economy. But seen from the side of managing payment instrument alternatives, one can see a clear interest in maintaining high volumes of cash and coins.

There are also some anomalies in the legislations of most European countries which favour cash and hinder automation (e.g. Germany, where the Federal Court27 forbids fees for cash withdrawal at a counter, but allows ATM fees – thus privileging the less automated variant and encouraging the use of cash).

Finally, national governments have benefited from the introduction of the euro notes and coins, since many people tend not to hand in their old currencies. In the Netherlands, only 50% of the total produced ‘guilder’ coins since 1948 were cashed in and changed to euros and cents.28 Nine years after the introduction of the euro as valid tender, the German Bundesbank says that DM13.45bn are still outstanding.29 These are again significant benefits to national governments.


A common understanding in the payments business is that electronic alternatives are faster and cheaper than manual, paper-based instruments. Australian, Belgian and Dutch central bank studies, however, found that the tender time at the point of sale was significantly less for a cash payment than for a payment card.30 Trained tellers, knowing what bill to expect, count out the change before customers have opened their wallets, whereas card payments involve remembering and entering PINs, getting the right cards the right way round, awaiting network connection and authorisation, perhaps signatures/PINs etc. Scott Thomson, director of QPQ, a payments adviser to retailers, showed in an empirical study that the difference in time to check out a customer paying by cash compared with another paying by card is at best ‘marginal’. Figure 4, also based on real empirical evidence,31 shows that cash is typically the fastest means of payment (until the advent of the contactless card).

Figure 4: The Average Duration of a Purchase Transaction by Payment Methods, in Seconds – Merchant Perspective



Knowing that large retailers consider that each second saved at the checkout adds another million euros in revenue per year, one can see that the speed of cash replacements can indeed have significant economic advantage to the retail industry: contactless cards save 8-12 seconds over cash per transaction.

Returning to cash and now looking not at speed but at comparative costs, the UK QPQ32 study asserts that, in highly optimised large retail stores, cash is not only the fastest but also the cheapest payment method: £0.02 per cash transaction versus £0.08 per debit card transaction (not to mention the cost of credit cards). In order to understand this, one may look at the large supermarket chains who are able to optimise their in-house cash logistics by, for example, feeding money from tills into in-house ATMs, providing cash-payout services at tills, having automated cash and coin-counting sorting machines for optimised end-of-day reconciliation. This trend is spreading to other industries, e.g. petrol stations which are installing their own cash recycling mechanisms.

Large chains such as Metro actually explicitly encourage customers to bring in coins. Thus their counting machines are reused; the customer shows the receipt of the count at the checkout. Not only does this attract customers and make cash hand -ling easier, Metro also charges a fee for helping merchants to keep their small change topped up (using coins gathered) and thus not only saves cost on cash, but also generates income.

Even smaller retailers (e.g. corner shops) appear more than willing to accept cash, both to meet customer demand and because they believe cash is cheaper than other payment methods. Signs such as ‘Card payment only above €20’ are ubiquitous. Even those retailers who have announced plans to stop accepting cash have done so for reasons such as security rather than cost and efficiency.

All retailers, large and small, suffer from expenses when using cards, owing to terminals (currently TA 7.0, EMV, PCI, etc), updates of standards on terminals, bank/card charges, bad debt and fraud.33 UK retailers thus argue that cash has ‘easily the lowest collection cost of any payment method’ (see Figure 5).

Figure 5: Collection Costs per Payment Instrument of Large UK Retailers

Source: RBR Analysis


It should, of course, be noted that much of the ‘cheapness’ of cash is artificial and due to its subsidisation.

It will be shown below that real value – competing successfully even against artificially subsidised paper instruments – can be gained by automating complete integrated value chains (e.g. e-payment after self-scanning in retail sales, or e-payment after e-invoice in online and corporate sales). Just attempting to optimise a local individual problem (card instead of cash at checkout) will yield only limited success compared with a complete solution.

It is also shown that new technologies such as biometric payments and contactless payments will finally prove to be the unchallenged champions, in particular addressing the issues of checkout speed and convenience, thus invalidating many of the concerns above. This, together with integration with loyalty systems will open new and very important avenues, which can finally lead to significant inroads against cash.


Finally, examining the stakeholder group of consumers, it is clear that they consider cash to be a winning proposition. In all criteria for judging a payment instrument, cash scores highly (see Figure 6). Not only is cash ‘free’ (since the actual societal costs of €130 per person are hidden and redistributed indirectly), it is accepted everywhere, users feel more in control, and it has become a strong habit.

Figure 6: Perceived Advantage by German Consumers of Various Retail Payment Instruments According to Main Criteria34

Source: Deutsche Bundesbank


Cash is also the only means of payment for those who do not have a bank account. This is a very large share of the developing world (in Ghana, 95% population are unbanked), but is also a significant market in the developed economies. For example, about five million people in the UK (10% of the population) and ten million households in the US (30%) have no bank accounts and thus can only resort to cash.35

Worldwide, there are over two billion people for whom cash is the only means of payment,36 and for most others cash is preferred for very understandable subjective and objective reasons. For the unbanked population, new means of money transfer such as mobile payments (m-payments) (see, for example, M-PESA in Kenya, where 6.5 million people have mobile phones but no bank accounts) are providing a very successful cashless solution – to the chagrin of banks, totally outside the financial services industry.

The local banks have been quick to make up lost ground however, now offering a bank-based savings solution (MKesha) on the basis of the m-payment solution. More concern is rightly being voiced by the national regulator which now finds one third of the local economy being sustained by an unregulated payment system.

More importantly on the global scale is the observation that this success seems to be a singularity: the development of this solution everywhere, not even in neighbouring Tanzania has hardly been on the scale observed in Kenya. This is surprising, since billions of people in the world are unbanked, and thus – years after Kenya – this mobile payment solution should have spread rapidly around the world.

Many efforts are being made: MTN mobile money is being tried in Uganda, Ghana, Cameroon, Ivory Coast, Ruanda and Benin. AirTel Money, also known as ZapZap, impressed with a new distribution model for Coca Cola. In the Philippines – without much marketing and with a poor network of agents – one in 10 unbanked mobile money users already stores an average of US$31 in their wallet. From Afghanistan to Zambia mobile network operators in developing countries are launching mobile money services at a rapid pace: to date 78 deployments have been launched and another 83 are being planned37. However, none of these are anything like the success of M-PESA (see Figure 7).38

Figure 7: Mobile Money Revenue in the Third World Compared with Expectations

Source: GSMA


For the developed world, this third world solution would clearly never be an option (although, even in the US, 56 million people do not have traditional bank accounts39), since regulation and demands on security are at a totally different level from sub-Saharan Africa. Indeed, for the developed world, a mobile solution which has a business case for banks is proving even more elusive, as discussed below.

But cash is also extremely successful in well-developed economies. Two countries whose cash holdings per capita are particularly high are Switzerland (Swiss francs equivalent to US$5245) and Japan (yen equivalent to US$6617).40 The extraordinarily high prevalence of cash in Japan can be explained by the safe environment in everyday life, the common distrust of banks based upon recent history, the fact that deflation actually encourages keeping cash (since its value, unlike in inflationary Europe, increases over time and thus discourages spending) and also by the habit of rich individuals of keeping large stores of cash well secured in their home to avoid capital gains tax and inheritance tax. The advantages to the Yakusa (Japanese mafia) are as outlined above. Japan’s highest denomination bank note, however, is the JPY10,000 bill (about €82) – much less compact than Europe’s €500 bill, so the Japanese Mafia do not enjoy quite the ‘privileges’ accorded to European criminals.

Japan is, of course, one of the most developed and technologically aware countries in the world, so the success (or lack thereof) of Japanese payment innovations as possible cash substitutes is discussed below.

In America – a geography very heavy on paper-based payments, more on cheques than cash – some inroads against cash are being made. Predictions41 are that Americans’ use of cash will decline by 4% a year between 2010 and 2015. Interestingly, some (30%) of consumers are using less cash than they did ,two years ago, while others (20%) claim to be using more. Even more surprisingly, the young ‘Generation Y’ (born from the mid-1970s on) – who are often particularly associated with new technology – are actually the only generation using old-fashioned notes and coins more than they did in 2008.

Broadly summarising the arguments in this section, there are many, and significant, stakeholders who have large interests in cash. There are huge economic advantages, some speed advantages, strongly emotionally felt advantages. Can these be overcome to provide a less paper-based, more efficient, less illegal economy?

Critical Success Factors

Cash provides many objective and subjective advantages to all stakeholders. In order to determine what the ‘best’ alternatives by electronic means would be, these should be measured against those factors that currently speak for cash.

Some factors spring quickly to mind: a retail payment instrument should be easy to use, readily available and accepted, should impose no prohibitive financial burden on the merchant and user, and should offer an appropriate level of security.

Even with these uncontentious criteria, one can already see a potential conflict of interests: security can usually only be realised at the price of added complexity. In the internet payment space, a common request of ‘a one-click payment which is totally secure’ will be unattainable. So even in these simple criteria one can already see that a balance or judged trade-off must be made between conflicting interests.

To explore this example further: if the environment is for low value payments, the security criterion can be relaxed (it must simply be ‘good enough’). By contrast, in the context of B2B treasury payments, only the highest levels of security are acceptable – which means that payments will only be authenticated with appropriately heavy mechanisms (e.g. multiple signatures, secure chip cards, multifactor authentication), which will necessarily reduce the ease of use. Thus, one sees that not only are there conflicts between criteria, but also the importance of the criteria themselves depends upon the usage scenario.

In an analysis by the European Payment Council (EPC) in preparation for selecting the ‘best answer’ from banks for e-commerce retail payments (currently dominated by solutions from non-traditional third parties such as PayPal), the criteria shown in Figure 8 were deemed relevant.42

Figure 8: EPC Criteria for Selection of Online Payment Instruments

Source: EPC


The criteria of importance depend heavily on the stakeholder view. Only the minority of criteria are common to all stakeholders and, as seen above, already these can cause conflicts. In general, each stakeholder (user, merchant, bank) will have different – possibly conflicting – interests and hence conflicting criteria with which to measure the quality of a new payment instrument.

Clear conflicts are seen between the end user (who would prefer to have the payment free) and a bank (who wishes to derive some benefit from providing a convenient service). An example of a conflict of interest between banks and merchants is currently reflected in the MIF discussions. An example of a conflict between merchant and end user would be whether the payment can be reversed: the user will wish for a possibility of refund, whereas the merchant wishes for a guaranteed, irrevocable payment, especially if the virtual goods (film, software download) the merchant has delivered are immediately and irrevocably consumed by the end customer after payment. Thus, all stakeholders in the relationship are in a conflict of interests, which will lead to different preferred choices and weightings of payment criteria.

Thus, choosing a new payment instrument is not a matter of taking a few criteria and optimising them. Instead, one is faced with a wealth of criteria which must be tuned to balance the interests of the conflicting parties. This is in every case a delicate business decision.

To exacerbate the matter further, the optimal criteria not only depend on which view one takes from which stakeholder, but also on the usage scenario. It has been shown above that the ‘security versus ease of use’ trade-off depends on the value of the transaction. More in-depth investigation will also show further trade-offs whose decision point depends upon the usage scenario. For example, ‘anonymity versus fraud detection’, where certain content will make it a priority for users not to identify themselves, but it will be necessary to supply (electronic or physical) addresses for delivery, and identification may also be necessary for fraud prevention and money laundering purposes.

In summary:

  • Many attributes are relevant (easy, secure, cheap).
  • They cannot be statically prioritised (order differs on usage scenario).
  • Attributes within user group conflict (e.g. secure versus easy to use).
  • Attributes between user groups conflict (e.g. consumer versus merchant demands on guarantee and anonymity).

This analysis leads to the understanding that a single payment solution (for example for all e-commerce) will not be sufficient. Depending upon scenarios, such as weights given to stakeholder concerns, different payment instruments will result. This may explain why there are currently over 300 different e-payment solutions in Europe.43

This number of parallel solutions is clearly wasteful of investments, provides no critical mass and is confusing to the end-users and merchants. Instead, it is felt that a ‘handful’ of e-payment solutions will be sufficient to cover a large section of requirements and markets.

Examples would be: