## Falling Use of Cash and Demand for Retail Central Bank Digital Currency — Introduction and Cash Use

## Source details

**Canonical URL:** [Falling Use of Cash and Demand for Retail Central Bank Digital Currency — Introduction and Cash Use](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022027-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022027-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022027-print-pdf.pdf.json)

---

### Overview
- Cash is demanded as a medium of exchange for payments, to hold as a precautionary reserve, and to facilitate illegal activities.
- Standard stock indicator: ratio of all coin and currency in circulation (CIC) to GDP.
- Prior findings preserved:
  - Ashworth and Goodhart (2020): CIC/GDP fell from its highest use in the late 1940s, flattened in the 1980s, and rose steadily after the early 2000s for the Euro Area, Japan, the UK and US.
  - Bech, et al. (2018): CIC/GDP rose for 20 out of 25 countries studied; it fell in 5 countries (China, Norway, Russia, South Africa, and Sweden). They concluded overall substitution of cards for cash was limited because about 75 percent of countries experienced an increase in CIC/GDP.
- CIC/GDP is a stock measure that can rise even when point-of-sale cash use falls, because CIC/GDP compares cash stock to the value of all payments used in GDP.
- Purpose: provide a focused market measure of cash use in retail/point-of-sale contexts, determine drivers of falling cash use, and assess how retail CBDC might affect market share of existing cash substitutes.

### Market-share measure of cash use (definition and rationale)
- Definition: CASH share = CASH/(CASH + CARDS + E-MONEY), where CASH is the value of cash withdrawn from ATMs.
- Rationale:
  - Focuses on flows used where cash normally circulates (retail, point-of-sale).
  - Indicates which substitutes (cards, e-money) have most affected cash use over time.
  - Avoids conflating cash stock growth with retail payment substitution dynamics.
- Data limitation noted: the CASH measure includes ATM withdrawals only; over-the-counter (OTC) cash withdrawals at financial institutions are missing for all but 6 of the 25 countries covered.

### Comparative measures and headline findings (2012–2019 coverage)
- Measures compared:
  - CIC/GDP (standard stock measure).
  - Small-value CIC/household consumption (HC).
  - CASH/(CASH + CARDS + E-MONEY) (flow-based market share).
- Cross-country results (2012–2019):
  - Cash use is falling in 8 countries using CIC/GDP.
  - Cash use is falling in 13 countries using small-value CIC/HC.
  - Cash use is falling in 24 countries using the CASH market-share measure.
- Observed patterns:
  - Countries with the lowest cash shares in 2019 (Korea, Sweden, and the U.S.) experienced relatively small reductions in their cash share over 2012–2019.
  - Countries with higher cash shares that experienced larger reductions over 2012–2019 include China, Spain, and Argentina.
  - Declines are generally slowing over time for many countries (falling at a decreasing rate).

### Key definitions and measurement details
- Small CIC:
  - Defined as currency denominations lower than the most supplied denomination in a country's ATM network.
  - For cross-country comparability the authors used $35 and under to define small CIC for all countries; multiplying $35 by the average PPP over 2012–2019 produced local currency cutoffs; if this cutoff matched the largest denomination the cutoff was reduced to the next lowest denomination.
  - Amromin and Chakravorti (2009) exchange-rate-adjusted average dollar value (excluding Switzerland) was $37 in their 13 OECD country sample.
- E-money: stored value arrangements (funds stored on a card or mobile phone, excluding limited or single-purpose e-money).
- Small CIC/HC: small CIC related to household consumption (HC) rather than GDP.

### Cross-country heterogeneity and notable country cases
- General finding: focusing on retail market flows increases the number of countries found to be reducing their relative use of cash; using the cash share measure reversed rising CIC/GDP and rising small CIC/HC measures to falling shares of cash use for the sample (except Euro Area and Hong Kong where data was unavailable).
- Country evidence and events:
  - China: 2019 payment survey — cash payments 16 percent; cards 23 percent; mobile payments 59 percent. Some 46 percent of respondents reported using no cash at all during the survey period. M0 rose from RMB 6.83 trillion by year-end 2016 to RMB 8.43 trillion by year-end 2020, an average annual increase of close to 6 percent.
  - Brazil: cash share fell by 22 pp in 2015; GDP fell by 3.5 percent in 2015; central bank created Fast Payment system (Pix); card expenditures rose by 9 percent in the crisis year.
  - India: demonetization in 2016 removed some 86 percent of currency; cash share fell by 17 pp over three years.
  - Russia: central bank promoted domestic card system (Mir), lowering interchange and offering incentives; policy aimed to reduce vulnerability to Western sanctions.
- Data caveats:
  - A proportion of currency values estimated to be held by other countries reduces reported CIC/GDP (Euro: 30 percent; US dollar: 55 percent).
  - OTC withdrawals and business cash withdrawals for making change are often not separable; including OTC data for 6 European countries would reduce the measured fall in cash share by varying pp amounts (Germany 7 pp, Spain 6 pp, Italy 2 pp, France 1 pp, Belgium & Netherlands <1 pp).
  - Some check, giro, or credit transfer payments may substitute for cash but instrument-level values are not available.

### Country cash-share levels in 2019 and annual change (selected table entries preserved exactly)
- Jurisdiction — Cash Share Level 2019 (%) — Annual Absolute Reduction (pp) — Annual Percent Reduction (%)
  - Korea — 6 — -0.5 — -5.0
  - Sweden — 8 — -1.3 — -7.5
  - United States — 10 — -0.5 — -4.0
  - Japan — 11 — -1.6 — -7.2
  - Australia — 16 — -1.5 — -5.7
  - United Kingdom — 16 — -1.6 — -5.8
  - France — 19 — -0.6 — -2.7
  - Netherlands — 20 — -1.9 — -5.7
  - Belgium — 25 — -2.7 — -6.1
  - Russia — 29 — -5.8 — -8.3
  - Switzerland — 31 — -0.1 — -0.3
  - China — 32 — -2.0 — -4.4
  - Singapore — 36 — -1.3 — -2.8
  - Spain — 36 — -2.0 — -3.9
  - Italy — 41 — -0.9 — -1.9
  - Brazil — 44 — -5.2 — -6.5
  - South Africa — 45 — -0.8 — -1.6
  - Argentina — 47 — -2.9 — -4.3
  - Turkey — 49 — 0.0 — 0.0
  - Canada — 49 — -0.2 — -0.3
  - Germany — 52 — -1.8 — -2.8
  - Mexico — 57 — -1.4 — -2.0
  - India — 64 — -3.6 — -4.0
  - Saudi Arabia — 72 — -1.6 — -1.9
  - Indonesia — 77 — -0.7 — -0.9
  - Euro Area — na — na — na
- Interpretation preserved:
  - Eight countries had cash share ranging from 6 percent to 20 percent in 2019 (Korea, Sweden, US, Japan, Australia, UK, France, Netherlands) with average reduction in cash share level of 1.2 pp a year.
  - At that rate it may take 5 to 17 years for the cash share measure to be close to zero.
  - The overall average percentage point reduction in the cash use measure across countries in Table 2 is 1.7 pp a year ((2012 cash share level minus 2019 level)/7).

### Time-series patterns and projection method
- Methodology:
  - Cash share levels plotted for 2012 (Y-axis) and 2019 (X-axis); a three-year projection to 2022 (sans Covid) was generated by estimating an inverse logistic curve ln[(1 - cash share)/cash share] = a + b x YEAR1 over 2012–2019 and predicting YEAR2 = 2020, 2021, 2022; a cubic spline was applied to the augmented sample to produce curves.
- Main takeaway: majority of cash share reductions are steady and smooth over time rather than volatile year-to-year.
- Largest absolute reductions: Brazil, India, Russia.

### Drivers: revealed preferences from payment diaries and demographic explanation
- Approach: use revealed preferences from payment diary studies rather than separately estimating many influences on payment choice.
- Key behavioral summary: younger adults reveal weaker preference for cash than older adults; demographic change favors substitutes.
- Simple demographic preference indicator: sumBRDR = cumulative sum over years of (birth rate + death rate), approximating annual change in intergenerational structure and revealed preference shifts.
- Illustrative payment-diary evidence (US, 2017–2019):
  - Total number of cash transactions: younger adults (18–44) = 3,119; older adults (45+) = 6,684.
  - Total value spent using cash: younger group = $94,541; older group = $153,988 (63 percent higher).
  - Population shares: older group represents 41 percent of US population; younger group represents 33 percent.

### Specified money-demand panel regression (Equation preserved)
- CASHSHAREj = β0 + β1*sumBRDRj + β2*HCj + β3*iRATEj + Vj
  - CASHSHARE = cash share values for 25 countries observed over 2012–2019 (N = 200).
  - sumBRDR = cumulative sum of birth rate + death rate for each country over 8 years.
  - HC = local currency PPP-adjusted US dollar value of household consumption.
  - iRATE = monetary policy related percent interest rate.
  - Estimation: Fixed Effects (Within) model with cross-country intercept dummies; also OLS on Means (Between).

### Pooled regression results (Table 3) — parameters and statistics preserved exactly
- Fixed Effects (Within):
  - Intercept: parameter = -.225; t-Statistic = -1.38
  - Sum of Birth Rate + Death Rate: parameter = -.773; t-Statistic = -11.94
  - Household Consumption: parameter = .305E-14; t-Statistic = .34
  - Interest Rate: parameter = .259E-03; t-Statistic = .33
  - Adjusted R2 = .97
- OLS on Means (Between):
  - Sum of Birth Rate + Death Rate: parameter = 6.28; t-Statistic = 3.92
  - Household Consumption: parameter = -.155E-13; t-Statistic = -1.26
  - Interest Rate: parameter = .234E-02; t-Statistic = .41
  - Adjusted R2 = .46
- Inference preserved:
  - Intergenerational change in population age structure (sumBRDR) is negatively and significantly related to within-country variation in cash share and explains a substantial portion of that variation.
  - Time-series changes in HC and iRATE are statistically insignificant in the Fixed Effects (Within) results.
  - Between-country shifts in sumBRDR are also statistically significant.

### Cost incentives to adopt retail CBDC — summary and channels
- Central observation: If retail CBDC is issued and structured to match card attributes (convenience, speed, fraud control), its adoption will depend on additional incentives relative to cards.
- Transaction-fee and cost considerations emphasized as main determinants: user fees, merchant interchange fees, merchant working capital costs.

A. Incentives for the payment user
- CBDC adoption depends on:
  - Matching or exceeding card benefits (convenience, speed, fraud control).
  - Lower or eliminated user fees relative to current payment instruments.
- User-fee specifics preserved:
  - Cash: inconvenient to obtain but typically free; opportunity cost is interest on deposits.
  - Cards: typically no user fee; retailers pay interchange fees instead of users.
  - Retail CBDC may have an explicit user fee, no fee, or may pay interest; treatment may differ across countries.
  - If CBDC has a user fee, that would favor card use; if CBDC has no explicit user fee (like cash), that incentive goes away.
- Privacy/anonymity:
  - Anonymity applies to cash; card purchase information is not truly anonymous.
  - Some CBDC transactions could be made anonymous by restricting value stored on a card or mobile phone for a sub-group of users.

B. Incentives for the payment receiver (merchants)
- Cost of accepting payments:
  - Receivers incur costs to accept cash (internal handling/monitoring, vault expense, insurance, armored couriers) and cards/e-money (interchange fees, terminal and communication link expense).
  - CBDC benefits and costs would likely be similar to cards but possibly without interchange fees.
  - Past studies usually show credit card costs exceed cash costs; evidence mixed for debit cards.
  - Merchants may be amenable to accepting an instrument without interchange fees.
- Surcharging and discounts:
  - In some countries, receivers can surcharge card users or offer discounts for cash; few do so unilaterally because of fears of lost sales and organized action often restricted by antitrust laws.
- Quick access to funds received:
  - Immediate finality with CBDC: funds received would be immediately deposited in the receiving firm's deposit account and available for immediate reuse.
  - Cash reuse requires deposit (one- or two-day lag); card funds availability typically one to three days; e-money backed by card similar.
  - Immediate availability can reduce receivers' working capital costs.
- Illustrative US example preserved:
  - US retail sales in 2018 = $5.27 trillion paid by cards with average interchange fee (and other expenses) of 1.5 percent.
  - Receivers would have paid $79 billion in fees.
  - If all retail sales were replaced by retail CBDC (or Fast Payments), receivers may save the card fees and save one to three days waiting for sales revenues.
  - With immediate availability of sales revenues, receiver working capital costs could be reduced by $505 million if one day is saved.
  - User fees for retail CBDC (or Fast Payments) may be charged in the US (due to the Monetary Control Act of 1980) and may be charged in other countries as well to recover costs.
  - Full advantage of saving card interchange fees may be weaker in countries that allow receivers to surcharge for card use.
- Possible card-firm responses:
  - Lower interchange fees and raise prices on other banking services to make up revenue losses.
  - Provide immediate availability of funds via jointly-owned special purpose deposit accounts to settle transactions immediately or with a slight delay.
  - Rely on Fast Payment services they own instead of competing with retail CBDC.

### Overall interpretation and policy implications (preserved)
- CIC/GDP stock measure can obscure retail substitution away from cash because many payments included in GDP do not compete with cash.
- Flow-based market-share measure CASH/(CASH + CARDS + E-MONEY) better captures retail cash use and substitution dynamics.
- Using the market-share measure, cash use declined in 24 of 25 countries over 2012–2019, indicating widespread retail substitution to cards and e-money.
- Decline in cash use is uneven across countries and over time; many countries show falling cash use at a decreasing rate.
- Changing age structure and revealed preferences from payment diaries provide a parsimonious explanation for cross-country declines in cash shares.
- If retail CBDC is issued and matches card attributes, incremental adoption will likely depend on transaction cost features: user fees, merchant interchange fees, and merchant working capital costs.
- Main short- to medium-term effect of CBDC introduction is likely to be on the share of current cash substitutes (primarily cards) rather than on cash itself, especially in countries where cards dominate the payment market (20 out of 25 countries in the sample).
- Recent annual average reduction in the share of cash use across countries: 1.7 pp.

### Additional preserved items and data notes
- Repeated textual item preserved: "0.054 in 2019." and "wpiea2022027-print-pdf - 0.054 in 2019." appear in the source.
- Appendix: data sources listed include BIS, ECB, IMF IFS, World Bank WDI, and central bank payment statistics for ATM cash withdrawals, bank notes in circulation, card payments, electronic money, GDP, household consumption, and interest rates.

*IMF Working Paper: Falling Use of Cash and Demand for Retail Central Bank Digital Currency — Introduction and Cash Use (wpiea2022027-print-pdf).*

### INTRODUCTION _____________________________________________________________________ 3

### INTRODUCTION

### Overview
- Cash is demanded as a medium of exchange for payments, to hold as a precautionary reserve, and to facilitate illegal activities.
- Standard indicator: ratio of all coin and currency in circulation (CIC) to GDP. Prior findings:
  - Ashworth and Goodhart (2020): CIC/GDP fell from its highest use in the late 1940s, flattened in the 1980s, and rose steadily after the early 2000s for the Euro Area, Japan, the UK and US.
  - Bech, et al. (2018): CIC/GDP rose for 20 out of 25 countries studied; it fell in 5 countries (China, Norway, Russia, South Africa, and Sweden). They concluded overall substitution of cards for cash was limited because about 75 percent of countries experienced an increase in CIC/GDP.
- The CIC/GDP ratio is a stock measure that can rise even when point-of-sale cash use falls, because CIC/GDP compares cash stock to the value of all payments used in GDP.

### Purpose and questions addressed
- Provide a focused market measure of cash use where cash normally competes: point-of-sale and retail contexts.
- Determine what is driving falling cash use across countries.
- Assess how retail CBDC, if issued, might affect the market share of existing cash substitutes (primarily cards) rather than the CIC/GDP stock measure.

*IMF Working Paper: Falling Use of Cash and Demand for Retail Central Bank Digital Currency — Introduction section.*

### Structure (as used in the paper)
- Section II: Adopt a market share measure of cash use and contrast it with CIC/GDP and small-value CIC/household consumption (HC).
- Section III: Explain cross-country declines in cash share using revealed preferences from payment diary studies and changing age structure.
- Section IV: Discuss implications for retail CBDC adoption given existing levels of cash substitution and card dominance.
- Section V: Summary. Appendix: Data sources.

---

### Market-share measure of cash use (definition and rationale)
- Defined measure: CASH share = CASH/(CASH + CARDS + E-MONEY), where CASH is the value of cash withdrawn from ATMs.
- Rationale:
  - Focused on flows used where cash normally circulates (retail, point-of-sale).
  - Indicates which substitutes (cards, e-money) have most affected cash use over time.
  - Avoids conflating cash stock growth with retail payment substitution dynamics.

### Data limitations noted
- Over-the-counter (OTC) cash withdrawals at financial institutions are missing for all but 6 of the 25 countries covered; large OTC cash acquisition can affect the measured market share of cash.
- The CASH measure includes ATM withdrawals only; OTC data omission is highlighted in the text.

---

### CASH USE: DIFFERENT MEASURES FOR DIFFERENT PURPOSES

### Comparative measures and findings
- Three measures compared in the paper:
  - CIC/GDP (standard stock measure).
  - Small-value CIC/household consumption (HC) — more representative of cash's role in purchases.
  - CASH/(CASH + CARDS + E-MONEY) (flow-based market share for retail competition).
- Cross-country results (2012–2019 coverage):
  - Cash use is falling in:
    - 8 countries using CIC/GDP,
    - 13 countries using small-value CIC/HC,
    - 24 countries using the CASH market-share measure.
- Observed patterns:
  - Countries with the lowest cash shares in 2019 (Korea, Sweden, and the U.S.) experienced relatively small reductions in their cash share over 2012–2019.
  - Countries with higher cash shares that experienced larger reductions over 2012–2019 include China, Spain, and Argentina.
  - Trend characterization: downward trend in cash use is not linear; for many countries it is falling at a decreasing rate, suggesting diminishing numbers of remaining users switching away from cash.

### Example statistic preserved
- Bech, et al. (2018) reported that about 75 percent of covered countries experienced an increase in their CIC/GDP ratio.

---

### A. Cash Share Changes Across Countries

- The paper documents cross-country heterogeneity in cash substitution:
  - Some countries already have low cash use levels, indicating substantial prior substitution to cards and e-money.
  - Countries differ in both level of cash share in 2019 and the magnitude of its change over 2012–2019.
- Specific country patterns highlighted (2012–2019 comparisons):
  - Korea, Sweden, U.S.: lowest cash shares in 2019 and relatively small reductions over 2012–2019.
  - China, Spain, Argentina: higher cash shares but greater reductions over 2012–2019.

---

### SOURCE OF THE DEMAND FOR CASH VERSUS ITS SUBSTITUTES

- Analytical approach:
  - Rather than separately estimating many influences on payment choice (convenience, relative cost, infrastructure availability, trust in electronic payments), the paper uses revealed preferences from payment diary studies.
  - Summary measure: marginal revealed preference for cash substitutes by younger adults relative to marginal preference for cash by older adults, reflecting changing population age structure.
  - Rationale: the revealed preference encapsulates the composite effect of the multiple influences on payment choice.

---

### COST INCENTIVES TO ADOPT RETAIL CBDC

- Central observation:
  - If retail CBDC is issued and structured to match card attributes (convenience, speed, fraud control), its adoption will depend on additional incentives relative to cards.
- Focused incentives relate to transaction fees and costs:
  - User fees.
  - Merchant interchange fees.
  - Merchant working capital costs.

### A. Incentives for the Payment User
- CBDC adoption by users depends on:
  - Matching or exceeding card benefits (convenience, speed, fraud control).
  - Lower or eliminated user fees relative to current payment instruments.

### B. Incentives for the Payment Receiver
- Merchant adoption hinges on:
  - Lower merchant payment acceptance costs (e.g., interchange fees).
  - Impacts on merchant working capital costs through payment timing and settlement features.

---

### SUMMARY

- The CIC/GDP stock measure can obscure retail substitution away from cash because many payments included in GDP do not compete with cash.
- A flow-based market-share measure, CASH/(CASH + CARDS + E-MONEY), better captures retail cash use and substitution dynamics.
- Using the market-share measure, cash use declined in 24 of 25 countries over 2012–2019, indicating widespread retail substitution to cards and e-money.
- The decline in cash use is uneven across countries and over time; many countries show falling cash use at a decreasing rate.
- Changing age structure and revealed preferences from payment diaries provide a parsimonious explanation for cross-country declines in cash shares.
- If retail CBDC is issued and matches card attributes, its incremental adoption will likely depend on fee and cost incentives affecting users and merchants, notably user fees, merchant interchange fees, and merchant working capital costs.

*IMF Working Paper: Falling Use of Cash and Demand for Retail Central Bank Digital Currency — Introduction and Cash Use sections.*

### 0.054 in 2019.

### wpiea2022027-print-pdf - 0.054 in 2019.

### Measures and definitions of cash use
- CIC/GDP and small CIC:
  - Small CIC defined as currency denominations lower than the most supplied denomination in a country's ATM network; for the US the most supplied denomination was $20 so $10, $5, and $1 notes plus coins were small CIC.
  - For cross-country comparability the authors used $35 and under to define small CIC for all countries.
  - Multiplying $35 by the average purchasing power exchange rate (PPP) over 2012–2019 produced local currency cutoffs; if this cutoff matched the largest denomination the cutoff was reduced to the next lowest denomination.
  - Amromin and Chakravorti (2009) exchange-rate-adjusted average dollar value (excluding Switzerland) was $37 in their 13 OECD country sample.
- Cash share measure:
  - Defined as CASH/(CASH + CARDS + E-MONEY).
  - Components: ATM cash withdrawals (CASH), debit and credit card payments (CARDS), and value of e-money payments (E-MONEY).
  - E-money defined as stored value arrangements (funds stored on a card or mobile phone, excluding limited or single-purpose e-money).
- Relation of small CIC to household consumption:
  - Small CIC is related to household consumption (HC) rather than GDP, producing the ratio small CIC/HC.

### Cross-country patterns and notable country cases (2012–2019)
- General findings:
  - Focusing more narrowly on where cash is used (GDP, household consumption, or market for cash) increases the number of countries found to be reducing their relative use of cash.
  - All rising CIC/GDP ratios and rising small CIC/HC measures in the sample were reversed to falling shares of cash use using the cash share measure (except where data was unavailable for the Euro Area and Hong Kong).
- Specific country evidence and events:
  - China: People's Bank of China (2021) payment survey reported value of cash payments in 2019 was 16 percent of the total; cards 23 percent; mobile payments 59 percent. Some 46 percent of respondents reported using no cash at all during the survey period. M0 rose from RMB 6.83 trillion by year-end 2016 to RMB 8.43 trillion by year-end 2020, an average annual increase of close to 6 percent.
  - Brazil: cash share fell by 22 pp in 2015; GDP fell by 3.5 percent in 2015; central bank created Fast Payment system (Pix); card expenditures rose by 9 percent in the crisis year.
  - India: demonetization in 2016 removed some 86 percent of currency; cash share fell by 17 pp over three years.
  - Russia: central bank promoted domestic card system (Mir), lowering interchange and offering incentives; policy aimed to reduce vulnerability to Western sanctions.
- Data caveats:
  - A proportion of currency values (Euro: 30 percent; US dollar: 55 percent) estimated to be held by other countries reduces reported CIC/GDP values accordingly.
  - Over-the-counter (OTC) withdrawals and business cash withdrawals for making change are often not separable in available data; including OTC data for 6 European countries would reduce the measured fall in cash share by varying pp amounts (Germany 7 pp, Spain 6 pp, Italy 2 pp, France 1 pp, Belgium & Netherlands <1 pp).
  - Some check, giro, or credit transfer payments may substitute for cash but instrument-level values are not available.

### Country cash-share levels in 2019 and annual change (Table 2)
- Jurisdiction — Cash Share Level 2019 (%) — Annual Absolute Reduction (pp) — Annual Percent Reduction (%)
  - Korea — 6 — -0.5 — -5.0
  - Sweden — 8 — -1.3 — -7.5
  - United States — 10 — -0.5 — -4.0
  - Japan — 11 — -1.6 — -7.2
  - Australia — 16 — -1.5 — -5.7
  - United Kingdom — 16 — -1.6 — -5.8
  - France — 19 — -0.6 — -2.7
  - Netherlands — 20 — -1.9 — -5.7
  - Belgium — 25 — -2.7 — -6.1
  - Russia — 29 — -5.8 — -8.3
  - Switzerland — 31 — -0.1 — -0.3
  - China — 32 — -2.0 — -4.4
  - Singapore — 36 — -1.3 — -2.8
  - Spain — 36 — -2.0 — -3.9
  - Italy — 41 — -0.9 — -1.9
  - Brazil — 44 — -5.2 — -6.5
  - South Africa — 45 — -0.8 — -1.6
  - Argentina — 47 — -2.9 — -4.3
  - Turkey — 49 — 0.0 — 0.0
  - Canada — 49 — -0.2 — -0.3
  - Germany — 52 — -1.8 — -2.8
  - Mexico — 57 — -1.4 — -2.0
  - India — 64 — -3.6 — -4.0
  - Saudi Arabia — 72 — -1.6 — -1.9
  - Indonesia — 77 — -0.7 — -0.9
  - Euro Area — na — na — na
- Interpretation:
  - Eight countries had cash share ranging from 6 percent to 20 percent in 2019 (Korea, Sweden, US, Japan, Australia, UK, France, Netherlands) with average reduction in cash share level of 1.2 pp a year.
  - At that rate it may take 5 to 17 years for the cash share measure to be close to zero.
  - The overall average percentage point reduction in the cash use measure across countries in Table 2 is 1.7 pp a year ((2012 cash share level minus 2019 level)/7).

### Time-series patterns and projections
- Figure 1 methodology:
  - Cash share levels plotted for 2012 (Y-axis) and 2019 (X-axis); a three-year projection to 2022 (sans Covid) was generated by estimating an inverse logistic curve ln[(1 - cash share)/cash share] = a + b x YEAR1 over 2012–2019 and predicting YEAR2 = 2020, 2021, 2022; a cubic spline was applied to the augmented sample to produce curves.
  - Main takeaway: majority of cash share reductions are steady and smooth over time rather than volatile year-to-year.
- Largest absolute reductions in cash share: Brazil, India, Russia (each had far larger absolute reductions than typical country-years).

### Demographic explanation: age-structure, payment diaries, and pooled regression
- Rationale:
  - Payment diary studies show younger adults reveal weaker preference for cash than older adults; younger cohorts therefore favor cash substitutes.
  - Authors construct a simple demographic preference indicator: sumBRDR = cumulative sum over years of (birth rate + death rate), approximating annual change in intergenerational structure and resulting revealed preference shifts.
- Illustrative payment-diary evidence (US, 2017–2019):
  - Total number of cash transactions: younger adults (18–44) = 3,119; older adults (45+) = 6,684.
  - Total value spent using cash: younger group = $94,541; older group = $153,988 (63 percent higher).
  - Population shares: older group represents 41 percent of US population; younger group represents 33 percent.
- Specified money-demand panel regression (Equation (1)):
  - CASHSHAREj = β0 + β1*sumBRDRj + β2*HCj + β3*iRATEj + Vj
  - Variables:
    - CASHSHARE = cash share values for 25 countries observed over 2012–2019 (N = 200).
    - sumBRDR = cumulative sum of birth rate + death rate for each country over 8 years.
    - HC = local currency PPP-adjusted US dollar value of household consumption.
    - iRATE = monetary policy related percent interest rate.
  - Estimation: Fixed Effects (Within) model with cross-country intercept dummies; also OLS on Means (Between).
- Pooled regression results (Table 3):
  - Fixed Effects (Within):
    - Intercept: parameter = -.225; t-Statistic = -1.38
    - Sum of Birth Rate + Death Rate: parameter = -.773; t-Statistic = -11.94
    - Household Consumption: parameter = .305E-14; t-Statistic = .34
    - Interest Rate: parameter = .259E-03; t-Statistic = .33
    - Adjusted R2 = .97
  - OLS on Means (Between):
    - Sum of Birth Rate + Death Rate: parameter = 6.28; t-Statistic = 3.92
    - Household Consumption: parameter = -.155E-13; t-Statistic = -1.26
    - Interest Rate: parameter = .234E-02; t-Statistic = .41
    - Adjusted R2 = .46
- Inference from regression and figures:
  - The intergenerational change in population age structure (sumBRDR) is negatively and significantly related to within-country variation in cash share and explains a substantial portion of that variation.
  - Time-series changes in household consumption (HC) and interest rate (iRATE) are statistically insignificant in the Fixed Effects (Within) results.
  - Between-country shifts in sumBRDR are also statistically significant.

### Overall interpretation and implications
- Cash shares are falling at a generally slow pace; countries with the lowest cash shares tend to have the smallest reductions.
- The demographic transition—more younger adults entering the population and fewer older adults exiting—appears to be an important driver of falling cash shares across countries.
- Substitutes for cash (cards and e-money) already hold market shares in many countries from 80 percent to 94 percent, and the share of substitutes is likely only to rise even with residual habit formation among continuing cash users.

*Source: wpiea2022027-print-pdf - 0.054 in 2019.*

### Introduction discussing Section III, we do not try to determine the importance of the many separate influences

### wpiea2022027-print-pdf - Introduction discussing Section III, we do not try to determine the importance of the many separate influences

### Summary findings on cash use trends
- Documented falling use of cash for payments across 25 countries over 2012–2019.
- All but one of 25 countries covered (Turkey) experienced a decline in relative cash use; some reductions were large due to special circumstances (Brazil, India, and Russia).
- Our measure: share of the value of cash used (cash withdrawn from ATMs) relative to the market for cash, consisting of the value of cash, cards, and e-money.
- Eight countries: cash substitutes make up 80 to 94 percent of the payment market where cash is primarily used.
- Recent annual average reduction in the share of cash use across countries: 1.7 pp.
- Substitution away from cash generally slows as the cash share becomes lower; countries with a very low cash share in 2019 had relatively small percentage point reductions over 2012–2019.
- Revealed preference from payment diary studies: younger adults marginally prefer cash substitutes while older adults marginally prefer cash; age-structure changes are an important determinant of cash share variation across the pooled 25-country sample.
- Economic activity and interest rates across countries were insignificant in explaining cross-country variation in cash share.

### Cost incentives to adopt retail CBDC (overview)
- Further reductions in cash use are likely to be slow and relatively predictable absent CBDC.
- CBDC could change the pace of substitution; CBDC attributes may be similar to cash and cards.
- For most countries (20 out of 25, Table 2, Column 1) the share of cards dominates that for cash and thus CBDC is likely to affect cards more than cash.
- Countries where cash use is already very low would likely see CBDC affect cards more than cash.
- Central banks’ supply-side rationales for issuing retail CBDC (as outlined in Bank for International Settlements (2021)) overlap with card benefits except for three transaction cost/fee considerations: user fees, merchant interchange fees, and merchant working capital costs — viewed here as main determinants of CBDC adoption incentives.

### A. Incentives for the payment user
- User fees / convenience:
  - Cash: inconvenient to obtain but typically free; opportunity cost is interest on deposits.
  - Cards: typically no user fee; innovating FinTech firms may charge users for extra convenience services.
  - Card user-fee absence means retailers pay interchange fees instead of users.
  - Retail CBDC may have an explicit user fee, no fee, or may pay interest; treatment may differ across countries.
  - If CBDC has a user fee, that would favor card use; if CBDC has no explicit user fee (like cash), that incentive goes away.
- Possible loss / fraud:
  - All payment types face loss or fraud risks; fraud prevention efforts are applied across cash, cards, e-money, and would likely be applied to CBDC (currency design, tokenization, PINs, AI).
- Universal acceptance / privacy / anonymity:
  - Current cash and cash substitutes have near universal acceptance; same is expected for CBDC.
  - Anonymity applies to cash; card purchase information is not truly anonymous.
  - Some CBDC transactions could be made anonymous by restricting value stored on a card or mobile phone for a sub-group of users.

### B. Incentives for the payment receiver
- Cost of accepting payments:
  - Receivers incur substantial costs to accept cash (internal handling/monitoring, vault expense, insurance, armored couriers) and cards/e-money (interchange fees, terminal and communication link expense).
  - Benefits and costs of CBDC would likely be similar to cards but possibly without interchange fees.
  - Past studies usually show credit card costs exceed cash costs; evidence mixed for debit cards.
  - Merchants may be amenable to accepting an instrument without interchange fees (potential incentive for CBDC adoption).
- Surcharging and discounts:
  - In some countries, receivers can surcharge card users or offer discounts for cash; fears of lost sales mean few do so unilaterally; organized action is often restricted by antitrust laws.
- Quick access to funds received:
  - Immediate finality with CBDC: funds received would be immediately deposited in the receiving firm's deposit account and available for immediate reuse.
  - Cash reuse requires deposit (one- or two-day lag); card funds availability typically one to three days; e-money backed by card similar.
  - Immediate availability can reduce receivers' working capital costs.
- Illustrative US example:
  - Assume all US retail sales in 2018 ($5.27 trillion) paid by cards with average interchange fee (and other expenses) of 1.5 percent of transaction value.
  - Receivers would have paid $79 billion in fees.
  - If all retail sales were replaced by retail CBDC (or Fast Payments), receivers may save the card fees and save one to three days waiting for sales revenues.
  - With immediate availability of sales revenues, receiver working capital costs could be reduced by $505 million if one day is saved.
  - User fees for retail CBDC (or Fast Payments) may be charged in the US (due to the Monetary Control Act of 1980) and may be charged in other countries as well to recover costs.
  - Full advantage of saving card interchange fees may be weaker in countries that allow receivers to surcharge for card use.
- Possible responses from card firms:
  - Card firms may lower interchange fees and raise prices on other banking services to make up revenue losses.
  - Card firms could provide immediate availability of funds by contributing to a jointly-owned special purpose deposit account to settle transactions immediately or with a slight delay inside/outside of the central bank.
  - Alternatively, card firms that own Fast Payment services may rely on those services instead of competing with retail CBDC.

### Summary and policy implications emphasized in the introduction
- The study focuses on the demand side for retail CBDC and incentives for adoption rather than ranking many separate influences on payment choice.
- If CBDC is structured to match user and merchant benefits of cards (convenience, speed, fraud control), adoption incentives will hinge on transaction cost features: user fees, merchant interchange fees, and merchant working capital costs.
- Treating CBDC like cash (no user fee) and allowing merchants to save card fees by accepting CBDC would be strong incentives for adoption.
- Immediate finality and immediate availability of funds for receivers could reduce working capital costs and encourage CBDC acceptance.
- Card schemes may counteract CBDC incentives by modifying interchange fees, providing immediate access to merchant revenues, or expanding Fast Payment services.
- The main short- to medium-term effect of CBDC introduction is likely to be on the share of current cash substitutes (primarily cards) rather than on cash itself, especially in countries where cards dominate the payment market (20 out of 25 countries in the sample).

### Appendix: Data sources (as listed)
- ATM cash withdrawal: Bank for International Settlements; European Central Bank
- Bank notes in circulation: Bank for International Settlements
- Birth rate and death rate: World Bank, World Development Indicators
- Card payments: Bank for International Settlements; European Central Bank
- Currency in circulation: International Monetary Fund, International Financial Statistics
- Electronic money: Bank for International Settlements; European Central Bank
- Gross domestic product: International Monetary Fund, International Financial Statistics
- Household consumption: International Monetary Fund, International Financial Statistics
- Interest rates: International Monetary Fund, International Financial Statistics

*IMF WORKING PAPERS Falling Use of Cash and Demand for Retail Central Bank Digital Currency*

### References

### References

### Academic studies on cash demand, payment choice, and consumer behavior
- Amromin, G., S. Chakravorti (2009), "Whither Loose Change? The Diminishing Demand for Small-Denomination Currency", Journal of Money, Credit and Banking, Vol. 41, No. 2/3: 315-335. 
- Arango-Arango, C., and N. Suarez-Ariza (2020), "Digital Payments Adoption and the Demand for Cash: New International Evidence", Journal of Payments Strategy & Systems, Vol. 14, No. 4: 392-410. 
- Ashworth, J., and C. Goodhart (2020), "The Surprising Recovery of Currency Usage", International Journal of Central Banking, Vol. 16, No. 3, June: 239-277. 
- Bagnall, J, D. Bounie, P. K. Huynh, A. Kosse, T. Schmidt, S. Schuh, and H. Stix (2016), "Consumer Cash Usage: A Cross-Country Comparison with Payment Diary Survey Data", International Journal of Central Banking, Vol. 12, No. 4, December:1-62. 
- Chodorow-Reich, G., G. Gopinath, P. Mishra, and A. Narayanan (2020), "Cash and the Economy: Evidence from India's Demonetization", Quarterly Journal of Economics, Vol. 135, No. 1: 57-103.  
- Hayashi, F. (2021), "Cash or Debit Cards?  Payment Acceptance Costs for Merchants", Economic Review, Federal Reserve Bank of Kansas City, Vol. 106, No3: (in press). 
- Hayashi, F., and E. Klee. 2003, "Technology Adoption and Consumer Payments: Evidence from Survey Data", Review of Network Economics, Vol. 22, No. 2: 175-90. 
- Koulayev, S., M. Rysman, S. Schuh, and J. Stavins (2016), "Explaining Adoption and Use of Payment Instruments by U.S. Consumers", The Rand Journal of Economics, Vol. 47, No. 2, April: 293-325. 
- Stavins, J. (2017), "How Do Consumers Make Their Payment Choices", Federal Reserve Bank of Boston Research Data Reports, No. 17-1, May. 
- Stavins, J. (2002), "Effect of Consumer Characteristics on the Use of Payment Instruments", New England Economic Review, Q3: 19-31. 
- Taylor, V. (2006). “Trends in Retail Payments and Insights from Public Survey Results”, Bank of Canada Review, Spring, 25-36. 
- Van Hove, L. (2006), "Why Electronic Purses Should be Promoted", Banking and Information Technology (Universitat Regensburg, Germany), Vol. 2, June: 20-31. 

### Central bank research, CBDC, and payments system reports
- Bank of France (2020), "Central Bank Digital Currency", February.  https://publications.banque-france.fr/sites/default/files/media/2020/02/04/central-bank-digital-currency_cbdc_2020_02_03.pdf  
- Bank for International Settlements (2021), Central Bank Digital Currencies: User Needs and Adoption, Report No. 3.  
- Bech, M., and J. Hancock (2020), "Innovations in Payments", BIS Quarterly Review, March: 21-36.  
- Bech, M., U. Faruqui, F. Ougaard, and C. Picillo (2018), "Payments are A-Changing' but Cash Still Rules", BIS Quarterly Review, March: 67-80. 
- Bech, M., Y. Shimizu, and P. Wong (2017), "The Quest for Speed in Payments", BIS Quarterly Review, March: 57-68. 
- Bindseil, U. (2020), "Digital Euro and Transformation of Payments", Slide 10, European Central Bank, October 22. 
- European Central Bank (2017), "Estimation of Euro Currency in Circulation Outside the Euro Area", External Statistics Division, ECB-Public area, https://www.ecb.europa.eu/pub/pdf/other/estimating_eur_in_circulation_outside_the_euro_area-201704.en.pdf 
- European Central Bank (2020), ECB Payment Statistics, September (the Blue Book). 
- People's Bank of China (2021), Progress of Research & Development of E-CNY in China, (in English), Working Group on E-CNY Research & and Development of the People's Bank of China, July. 
- Swiss National Bank (2020), Survey on Payment Methods, Zurich.  
- Yoshizawa, K., K. Maehashi, H. Yanagihara, Y. Kadogawa, and M. Inada (2021), “Developments in Banknotes in Circulation since the Start of the Pandemic”, Bank of Japan Review, December. 
- Judson, R. (2017), "The Future of Cash in Crisis and Calm: Demand for US Dollar Banknotes", in War on Cash: Is There a Future for Cash? International Cash Conference 2017, Deutsche Bundesbank: 200-248. 

### Empirical analyses, cost studies, and payment diaries
- Allpago, Credit and Debit Card Purchases in Brazil Increase and Total R$ 1.05 Trillion in 2015, https://www.ppro.com/insight/credit-and-debit-card-purchases-in-brazil-increase-and-total-r-1-05-trillion-in-2015/(accessed September 17, 2021). 
- Food Marketing Institute (1998). A Retailers Guide to Electric Payment Systems Costs, Washington, D.C. 
- Federal Reserve Bank of Atlanta (2020), 2019 Diary of Consumer Payment Choice, data files are found at: https://www.atlantafed.org/banking-and-payments/consumer-payments/diary-of-consumer-payment-choice/2019-diary.aspx 
- Chen, H., W. Engert, M. Felt, K. Huynh, G. Nicholls, D. O’Habib, and J. Zhu (2021), “Cash and COVID-19: The Impact of the Second Wave in Canada”, Staff Discussion Paper 2021-12, Bank of Canada, July. 
- Hayashi, F., and J. Maniff (2021), "Public Authority Involvement in Payment Card Markets: Various Countries August 2021 Update". Federal Reserve Bank of Kansas City. 
- Humphrey, D., M. Kim, and B. Vale (2001), "Realizing the Gains from Electronic Payments: Costs, Pricing and Payment Choice", Journal of Money, Credit and Banking, Vol. 33, No. 2, May: 216-234. 
- Humphrey, D., and R. Hunt (2013), "Cost Savings from Check 21 Electronic Payment Legislation", Journal of Money, Credit and Banking, Vol. 45, No. 7, October: 1415-1429. 
- Kim, L., R. Kumar, and S. O'Brien (2020), 2020 Findings from the Diary of Consumer Payment Choice, Cash Product Office, Federal Reserve System, July. 
- Koulayev, S., M. Rysman, S. Schuh, and J. Stavins (2016), "Explaining Adoption and Use of Payment Instruments by U.S. Consumers", The Rand Journal of Economics, Vol. 47, No. 2, April: 293-325. 
- Paymnts.com (2018), Global Cash Index, United States Edition, March. 
- The Economist (2021), "Homegrown Fare", August 28: 63-64. 
- Khiaonarong, T., and D. Humphrey (2019), "Cash Use Across Countries and the Demand for Central Bank Digital Currency", Journal of Payments Strategy and Systems, Vol. 13, No. 1: 32-46. 

*Falling Use of Cash and Demand for Retail Central Bank Digital Currency — Working Paper No. WP/22/27*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022027-print-pdf.pdf_