Skip to main content

Advertisement

SpringerLink
Log in

Choosing and using payment instruments: evidence from German microdata

  • Published:
Empirical Economics Aims and scope Submit manuscript

Abstract

Germans are still very fond of using cash. Of all direct payments transactions in 2008, cash accounted for an astounding 82 % in terms of number and for 58 % in terms of value. With a dataset that combines transaction information with survey data on payment behaviour of German consumers, we shed light on how individuals decide on their cash usage. We employ a two-stage empirical framework which jointly explains payment card ownership and the use of cash. Our results indicate that cash usage is compatible with systematic economic decision making. Consumers decide on the adoption of payment cards and then use available payment media according to transaction characteristics, the relative costs of cash and card usage, socio-demographic characteristics and their assessment of payment instruments’ characteristics. Importantly, older consumers use significantly more cash than younger consumers. We show that this difference in payment behaviour is not attributable to age as such but largely to differences in the characteristics of older and younger consumers. This suggests that the high cash intensity of older consumers cannot fully be attributed to the role of habit or to their slow adoption to new payment technologies.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Notes

  1. Bundesbank survey “Payment habits in Germany”. The figures are very close to results for Austria where cash payments accounted for 86 % of all direct payment transactions in 2005 (Mooslechner et al. 2006). Direct payment transactions comprise all transactions apart from recurrent transactions, which are typically settled by direct debit or by bank transfers (e.g. rents, insurance fees, telephone bills).

  2. Examples of cash intensive countries for which data are available are Germany, Spain, Italy and Austria.

  3. Data from 2009 show that the density of payment card accepting devices is still rather low in Germany (7,883 per million inhabitants), in particular in comparison to countries where cash has lost some ground, like the Netherlands (14,779) or Finland (32,995). This indicator alone, however, cannot fully explain international differences in cash usage because the density of payment terminals is also relatively high in countries in which cash has remained important, like Spain, Italy or Austria (Spain: 30,324, Italy: 24,233, Austria: 12,079). Source: http://www.ecb.int/paym/market/blue/html/index.en.html, all values from 2009.

  4. Some of these hypotheses are competing. For example, Markose and Loke (2003) argue that cash and card payments are perfect substitutes while Drehmann et al. (2002) maintain that cash and payment cards are not perfect substitutes because cash has the distinctive feature of preserving anonymity.

  5. In Germany overdraft credit lines of checking accounts are widespread, and people can access them using their debit card. Almost everybody pays off credit card balances in full at the end of the month. According to the ECB Blue Book as of February 2009 (ECB 2009), only 2.6 million credit cards in Germany are equipped with a credit function, while 11.6 million credit cards are only providing delayed debit functionality.

  6. Ideally, one would need to have individuals observed over time to identify the role of habits. Such data are not available.

  7. Note that we treat the technical payment infrastructure, such as the number of card payment terminals, as given. For example, Markose and Loke (2003) or Rysman (2007) focus on both the demand and the supply side.

  8. Schuh and Stavins (2010), which contains many interesting insight, do not focus on cash but on check usage in the U.S. Nevertheless, the paper reports a specification explaining the cash share of expenditures which does not explicitly account for relative costs of cash (like shoe leather costs).

  9. In contrast to Bounie and Francois (2006) and Hayashi and Klee (2003), for example, we do not have information on the physical characteristics of the point of sale (e.g. the absence of a cashier or the availability of self-service).

  10. Borzekowski and Kiser (2008) are the only example we are aware of. In particular, in a counterfactual exercise the population is “aged” and the authors analyse how this affects hypothetical market shares of various payment instruments in the U.S. In contrast to our approach, these market shares are not observed but only constructed from survey answers on the relative ranking of payment instruments. Moreover, market shares are constructed under the assumption that all survey respondents conduct the same number of transactions.

  11. In our estimations, we calculate a simulated likelihood on the basis of pseudo-random variates using the Geweke–Hajivassiliou–Keene (GHK) simulator with 2,000 draws.

  12. The sampling technique comprised three stages. In the first stage, regions were selected (“sample points”), which were used to define starting points/addresses for the second stage, in which interviewers contacted households based on a random route procedure. Finally, an eligible person in each contacted household was randomly selected.

  13. We only collect information on direct payment transactions in the analysis, i.e. all transactions apart from recurrent transactions, which are typically settled by direct debit or by bank transfers (e.g. rent, insurance fees, telephone bills, utility bills).

  14. There are just 165 persons not owning any cards. The implied lack of variation renders it difficult to implement a meaningful econometric model of the debit card adoption decision.

  15. In principle, the information about the cash share for different expenditure types could also be extracted from the short-run payment diary data. However, most of the transactions recorded in the diary are retail transactions (44 %) and no other spending place reaches more than 10 % of total transactions recorded. Thus, there is only a very small number of transactions other than retail. Given that we also exclude transactions where no alternative media of payment was accepted, the number would be even lower. Therefore, we resort to the long-run payment behaviour as described by the CAPI data.

  16. The survey questions underlying the preference variables refer to payment instruments in general and not to a particular payment instrument.

  17. When constructing P_ABROAD and P_INTERNET, we code the indicator as 1 if the respective quality is regarded as “indispensable” or “rather important”, due to the small number of respondents choosing the highest ranking.

  18. The validity of our instruments is confirmed by the Sargan test \((p = 0.41)\) and the low correlation between the residuals and the credit card dummy \((-0.06)\). We also perform a test of “weak instruments” and find a Shea’s \(R^{2}\) of 2.3 % and a Cragg–Donald Wald F statistic of 5.7, which is just above the critical values for the weak instrument test reported in Stock and Yogo (2005). Additionally, we perform three different hypothesis tests following Stock et al. (2002) that are robust to weak instruments. All three tests confirm that the credit card variable is insignificant when instrumented. We also re-estimate the model using Stata’s condivreg command, which allows robust inference in the presence of potentially weak instruments. The results do not change.

  19. As noted above, the LHS variable is the share based on the volume of transactions. The results for the share based on the value of transactions are very similar, qualitatively.

  20. That habit is not of outstanding importance for cash usage can also be inferred from looking at raw survey results. For 56 % for respondents, habit is not important in choosing a payment instrument. Among those who state that habit is important, only one third states that only cash fulfils this attribute. In other words, these results show, not surprisingly, that consumers are as well acquainted with cash as well as payment cards.

  21. We want to remind the reader that supply effects cannot be the source behind the importance of the type of transaction—we analyse only transactions for which consumers have a payment choice.

  22. Currently, electronic point-of-sale terminals used by merchants have the technology to process both debit cards and credit cards. However, there are transaction types, such as in grocery stores, where debit card payments are allowed but not credit card payments. Given the technical infrastructure, the opposite is less likely, as purely paper-based credit card payments are about to vanish. This could imply that the coefficient for the POS density could actually reflect past rather than current POS densities, when the technology gap between debit and credit card payments was larger.

  23. Not interacted are the three indicators for education status, the gender variable and the frequencies of expenditure for given transaction types included in the OLS and IV regressions.

  24. A comparison between the mean for older consumers and the mean for the full sample yields qualitatively the same results. The statistics for this mean comparison are available upon request.

  25. The null hypothesis that all interaction terms are jointly insignificant is rejected for the multivariate probit’s credit card equation \((p <0.01)\), the simple probit equation for credit card ownership \((p<0.01)\), the OLS regression for the cash share \((p < 0.05)\) and the multivariate probit’s daily retail equation \((p <0.05)\). The null is not rejected for the gas station equation \((p = 0.37)\) and the IV equations (p value: 0.15).

  26. For the share of cash payments, we use OLS estimates or this decomposition, as OLS is the best linear predictor.

  27. In addition, we vary the number of pseudo-random draws (100, 1,000, 2,000) and seeds for the multivariate probit. We also use different simulation methods (GHK, Halton draws).

  28. The test statistics are \(\chi ^{2}(32) = 55.60\) with \(p > \chi ^{2} = 0.006\) if we exclude the constant from the comparison and \(\chi ^{2}(33) = 248.31\) with \(p > \chi ^{2} = 0.000\) if we include the constant.

  29. Zinman (2009) draws a similar conclusion for the use of debit and credit cards.

  30. This could indicate that only one of them or a combination of both may survive in the long run. However, as credit cards have some distinctive features which debit cards currently do not typically have, like travel insurance coverage or the possibility to make payments abroad, it is unlikely that credit cards will disappear.

References

  • Alvarez FE, Lippi F (2009) Financial innovation and the transactions demand for cash. Econometrica 77(2):363–402

    Article  Google Scholar 

  • Arango C, Huynh KP, Sabetti L (2011) How do you pay? The role of incentives at the point-of-sale. Bank of Canada working paper no. 2011-23

  • Amromin G, Chakravorti S (2007) Debit card and cash usage: a cross-country analysis. Federal Reserve Bank of Chicago working paper 2007-04

  • Attanasio OP, Guiso L, Jappelli T (2002) The demand for money, financial innovation, and the welfare cost of inflation: an analysis with household data. J Political Econ 110(2):317–351

    Article  Google Scholar 

  • Baumol WJ (1952) The transaction demand for cash: an inventory theoretic approach. Q J Econ 66(3): 545–556

    Google Scholar 

  • Boeschoten WC (1998) Cash management, payment patterns and the demand for money. De Econ 146(1):117–142

    Google Scholar 

  • Borzekowski R, Kiser EK (2008) The choice at the checkout: quantifying demand across payment instruments. Int J Ind Organ 26:889–902

    Article  Google Scholar 

  • Borzekowski R, Kiser EK, Ahmed S (2008) Consumers’ use of debit cards: patterns, preferences, and price response. J Money Credit Bank 40:149–172

    Article  Google Scholar 

  • Bounie D, Francois A (2006) Cash, check or bank card? The effects of transaction characteristics on the use of payment instruments. SSRN eLibrary. http://ssrn.com/paper=891791. Accessed 19 Feb 2009

  • Drehmann M, Goodhart C, Krueger M (2002) The challenges facing currency usage: will the traditional transaction medium be able to resist competition from new technologies? Econ Policy 17(34):193–227

    Article  Google Scholar 

  • ECB (2009) Payments and settlement systems statistics. http://sdw.ecb.europa.eu/reports.do?node=1000001114. Accessed 19 Feb 2009

  • ECB (2011) The use of euro banknotes—results of two surveys among households and firms. European Central Bank Monthly Bulletin, Frankfurt, pp 79–89

  • Hayashi F, Klee E (2003) Technology adoption and consumer payment: evidence from survey data. Rev Netw Econ 2(2):175–189

    Article  Google Scholar 

  • Humphrey DB, Kim M, Vale B (2001) Realizing the gains from electronic payments: costs, pricing, and payment choice. J Money Credit Bank 33(2):216–34

    Article  Google Scholar 

  • Klee E (2008) How people pay: evidence from grocery store data. J Monetary Econ 55(3):526–541

    Article  Google Scholar 

  • Lippi F, Secchi A (2009) Technological change and the households’ demand for currency. J Monetary Econ 56(2):222–230

    Article  Google Scholar 

  • Mantel B (2000) Why do consumers pay bills electronically? An empirical analysis. Econ Perspect Fed Reserv Bank Chic Q4:32–48

    Google Scholar 

  • Markose SM, Loke YJ (2003) Network effects on cash card substitution in transactions and low interest rate regimes. Econ J 113(487):456–76

    Article  Google Scholar 

  • Mooslechner P, Stix H, Wagner K (2006) How are payments made in Austria? Results of a survey on the structure of Austrian households’ use of payment means in the context of monetary policy analysis. Monetary Policy Econ Q2:111–134

    Google Scholar 

  • Rysman M (2007) An empirical analysis of payment card usage. J Industrial Econ 55: 1–36

    Google Scholar 

  • Santomero AM, Seater JJ (1996) Alternative monies and the demand for media of exchange. J Money Credit Bank 28(4):942–60

    Article  Google Scholar 

  • Schuh S, Stavins J (2010) Why are (some) consumers (finally) writing fewer checks? The role of payment characteristics. J Bank Finance 34(8):1745–1758

    Article  Google Scholar 

  • Schuh S, Stavins J (2012) How consumers pay: adoption and use of payments. Federal Reserve Board Boston working paper series 12-2

  • Shy O, Tarkka J (2002) The market for electronic cash cards. J Money Credit Bank 34(2):299–314

    Article  Google Scholar 

  • Stix H (2004) How do debit cards affect cash demand? Survey data evidence. Empirica 31(2–3):93–115

    Article  Google Scholar 

  • Stock JH, Yogo M (2005) Testing for weak instruments in linear IV regression. In: Andrews DWK, Stock JH (eds) Identification and inference for econometric models: essays in honor of Thomas Rothenberg. Cambridge University Press, New York, pp 80–108

    Chapter  Google Scholar 

  • Stock JH, Wright JH, Yogo M (2002) A survey of weak instruments and weak identification in generalized method of moments. J Bus Econ Stat 20(4):518–529

    Article  Google Scholar 

  • Tobin J (1956) The interest elasticity of transactions demand for cash. Rev Econ Stat 38(3):241–247

    Article  Google Scholar 

  • Whitesell WC (1992) Deposit banks and the market for payment media. J Money Credit Bank 24(4):483–498

    Article  Google Scholar 

  • Zinman J (2009) Debit or credit? J Bank Finance 33(2):358–366

    Article  Google Scholar 

Download references

Acknowledgments

We thank two anonymous referees for very helpful comments. This paper represents the authors’ personal opinions and does not necessarily reflect the views of the Deutsche Bundesbank, the Oesterreichische Nationalbank or their staff.

Author information

Authors and Affiliations

  1. Deutsche Bundesbank, Frankfurt, Germany

    Ulf von Kalckreuth & Tobias Schmidt

  2. Economic Studies Division, Oesterreichische Nationalbank, POB 61, 1011 , Vienna, Austria

    Helmut Stix

Authors
  1. Ulf von Kalckreuth
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Tobias Schmidt
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Helmut Stix
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Helmut Stix.

Appendix

Appendix

See Tables 6 and 7.

Table 6 Construction of variables
Full size table
Table 7 Descriptive breakdown of payment behaviour indicators
Full size table

Rights and permissions

Reprints and permissions

About this article

Cite this article

von Kalckreuth, U., Schmidt, T. & Stix, H. Choosing and using payment instruments: evidence from German microdata. Empir Econ 46, 1019–1055 (2014). https://doi.org/10.1007/s00181-013-0708-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00181-013-0708-3

Keywords

JEL Classification

Search

Navigation