Abstract
Purpose
This paper aims to assess whether the coronavirus disease 2019 (COVID-19) pandemic has encouraged governments to take actions towards fostering digital means of payments and financial transactions to stimulate economic activities and achieve higher financial inclusion.
Design/methodology/approach
Using a logit model, this paper tests the impact of the level of income and GDP per capita, government effectiveness, digital adoption, number of commercial banks and the pandemic-related closure of business and stores due to full lockdowns on governments’ policy response regarding digital means of payments.
Findings
The author finds that low- and lower-middle-income countries had significantly responded to the surged need for digital means of payment during the pandemic compared to the upper-middle-income and high-income countries. The author also finds that government effectiveness and the number of commercial banks were predictors of government policy response, while the full lockdown of countries and the overall digital adoption were not.
Research limitations/implications
Data of the post-COVID-19 pandemic are limited, and the sample size is small.
Originality/value
This is the first paper to empirically model governments' response during the pandemic to promote digital means of payments. This paper gives insight into post-crisis potential changes in digital payment adoption in the upcoming years.
Keywords
Citation
Mansour, H. (2022), "How successful countries are in promoting digital transactions during COVID-19", Journal of Economic Studies, Vol. 49 No. 3, pp. 435-452. https://doi.org/10.1108/JES-10-2020-0489
Publisher
:Emerald Publishing Limited
Copyright © 2021, Emerald Publishing Limited
1. Introduction
In the wake of every global economic crisis, governments and central banks respond to contain any negative effect and to seize any potential opportunity. The global financial crisis that occurred in 2008 has pushed many countries which were affected by the crisis to take a new direction to support their digital financial infrastructure, which has resulted in the subsequent growth of the FinTech worldwide (Menat, 2016). The term “FinTech” was introduced in the early 90s and refers to delivering any financial solution through technology. The rapid spread and the fear of exposure to the coronavirus has forced governments to restrict social gatherings, close malls and reduce transportation capacity and close schools and universities.
As a response, many central banks took actions to maintain consumption and production by adopting policies to support their digital payment systems. Such policies included reducing costs of digital payments transactions, promoting digital distribution channels and mobile money, allowing digital banking documentation and extending limits of transactions. Governments' responses to coping with the “new normal” had allowed some countries to adjust their situations. Before the pandemic, there has been an increase in the adoption of digital online means of payments or mobile money; non-cash transaction volumes were globally steadily growing with emerging markets growing at a double rate of mature markets rates (Capgemini, 2019), and financial inclusion and digital payment values have sharply increased with millions of individuals counted in the formal economy. Nevertheless, some developing countries could stimulate digital payment mechanisms before the pandemic; however, it was not set as a priority. The pandemic can be the catalyst for digital and electronic payments and services to flourish.
Some countries are still in the very early stages in digitizing their payments. The financial digitalization started in the 60s when cards and automated teller machines (ATMs) were the first digitalized technologies to enter the banking system, and in the 80s, the Internet era has begun and connected buyers and retailers with their suppliers and sellers. Since then, big data analysis and artificial intelligence have led many financial businesses (Moşteanu et al., 2020). In this regard, many governments are trying to pave a sustainable path for digital payment systems by taking short- and long-term policies to adopt modern technologies. For countries that are still in earlier stages of financial digitization, innovative solutions can help in fostering active usage of banking services and digital means of payments to rely on them in day-to-day transactions. This will also facilitate the transfer of funds in times of hardships; the COVID-19 crisis was a clear example. To reach the e-payment maturity state experienced in many developed nations, a vision of sustainable digital inclusion and effective policy formulation is needed. That being said, this paper aims to assess whether the coronavirus disease 2019 (COVID-19) pandemic has encouraged low- and lower-income governments to take actions towards stimulating economic activities through the usage of digital means of payments and online financial transactions. Using a logit model, I test the impact of countries' income level, government effectiveness level, digital adoption, number of commercial banks and the pandemic-related closure of business and stores due to full lockdowns on governments' policy response to promote digital means of payments. In the next section, I present a literature review followed by a data and methods section and a final section for discussion and conclusion.
2. Literature review
The global economic and financial impact of the COVID-19 is becoming a rich research material for economists worldwide. A number of studies have discussed the COVID-19 impact on the use of digital transactions. Remolina (2020) find the pandemic as an opportunity for FinTechs because they are now viewed as a part of the solution for the pandemic-related challenges. In the long term, consumers' risks and financial sector competition and regulations will determine the sustainability of FinTechs. Dubey et al. (2020) discussed how digital complaint handling, digital onboarding and remittance transfers “with the speed of light” are some of the features that encouraged customers to use financial technologies during the pandemic; as a customer's reach was “exponentially growing” during the lockdown. Allam (2020) showed Information and communications technology (ICT) and payment corporations' capability and views the pandemic as a push towards cashless systems worldwide.
On the contrary, a study done by Chen and Zhu (2020) on survey data from Canada showed that cash in circulation has increased during the peak of the pandemic and that access to cash and cash holdings increased during that time compared to the use of e-transfers but less than the use of debit/ credit cards. Tut (2020) has also investigated the impact of the crisis on adopting both individuals and financial entities of FinTech as a means for payments in Kenya. The study concluded that the crisis had a negative impact on the use of FinTech but a positive one on the regulatory framework. Aji et al. (2020) examined the impact of “government intention to support digital solutions” as well as “perceived risk” and “usefulness” on the willingness of individuals of both countries to use e-wallets. The study was done on Malaysia and Indonesia. It showed that e-wallets supported by the government during the pandemic had different outcomes on individuals of both countries, and that the “perceived usefulness” was the mediator in determining the effect of government support. Besides, a study done by Strusani and Houngbonon (2020) discussed digital connectivity as a component of resilience for both businesses and communities because promoting the digital inclusion of low-income groups eases the transfer of funds for both individuals, small and micro-businesses during emergencies and over-recovery periods.
Digital systems open the opportunity to support greater financial inclusion and resilience for populations affected by the crisis. Another view of resilience was discussed by Bounie et al. (2020), who studied data from France and concluded that the consumption shock that happened in France was compensated by the online shopping shifting, which is a sign of resilient economy at time of contractions. Digital transactions can actually promote financial inclusion as societies' behaviour changes (Auer et al., 2020a, b).
Indeed, digital infrastructure is the key starting point if governments want to minimize the pandemic's economic costs through digital means and ensure the reception of these services by all targeted individuals (de Mello and Ter-Minassian, 2020). Thus, a successful government digital transformation response requires a clear and defined vision and goal, as the quality of intervention is more important than simply intervening (Schiliro, 2020).
Many factors explain the low adoption of e-payments in developing countries. Frost (2020) gives a more in-depth explanation and states that the low adoption results from demand for financial services that suppliers do not meet. In more developed economies, Frost (2020) suggests that the adoption level is determined by the cost of transactions and regulatory frameworks. Frost (2020) added that countries with a higher percentage of youth are more likely to deal with financial technologies. Sha'ban et al. (2020) showed how most countries in Africa and the South and East Mediterranean have high levels of financial exclusion and that Sub-Saharan African countries rank the lowest across all other regions. In Nigeria, Ayo and Babajide (2006) argued that ignorance, illiteracy and security issues are the determining factors; however, some psychological and cultural reasons behind the rejection of using e-payments; people want to touch and physically hold cash. Within the same discussion, Ligon et al. (2019) showed how digital transactions costs are unlikely to predict the adoption of digital transactions and that demand-side factors, such as social norms, are the key factors in India. Barriers to the adoption of electronic payments in low-income groups were discussed by Ellison et al. (2012) and were divided into (1) societal and environmental; (2) banking and financial exclusion; (3) payment infrastructure; (4) payment services systems and (5) money management practices.
Likewise, Augsburg and Hedman (2014) showed how mobile money use is very low in developing countries despite being popular as an emerging market. Indeed, a driving factor behind financial inclusion could be mobile money (Patil et al., 2017). The unified theory on acceptance and use of technology (UTAUT) is an interesting theory introduced to explain users' technology adoption behaviour in general. The theory combines around eight theories and distinguishes between direct determinants, such as(1) performance expectancy; (2) effort expectancy; (3) social influence and (4) facilitating conditions, and other mediators, such as (1) gender; (2) age; (3) experience and (4) voluntariness of use.
The fact is that the difference between immature and mature digital finance enabled markets in developed and developing countries emerges from developed countries' long process. Early stages of adoption entail consumer acceptance, ease of use and risk compared to traditional banking. In an advanced stage, convenience, speed, security and privacy become more important in determining the level of adoption (Singh, 2019).
In addition to that, a robust and efficient banking sector is a prerequisite for financial and digital inclusion from the supply side. Banks try to maximize profits and lower its costs, while concentration leads to lower competition between banks (Gharsellaoui, 2015). Such competition can impact the efficiency of the banking sector. However, many studies have linked the number of banks in an economy with cost and profit efficiency. In many of them, bank competition is believed to have a negative effect on cost and profit efficiency (Thom and Thuy, 2019). However, competition in many developing and emerging economies is very limited, and this lack of competition results in more costly and lower quality of financial services.
Many factors determine ICT adoption in general, and government quality is one of them. In a study on Sub-Saharan Africa countries, Asongu and Biekpe (2017) found that besides technical determinants of ICT, government effectiveness has a positive short-run and long-run impact on ICT adoption. Government effectiveness also positively impacts financial inclusion and can determine central banks' tendency to adopt a central bank's digital currency (Auer et al., 2020a, b).
As discussed, many developing countries lag in financial and digital inclusion; however, the pandemic had given some countries a push towards higher adoption and promotion of digital financial payments and services. In the next section, I present data and methods to evaluate the response regarding e-payments of different countries with different income levels, digital adoption levels and government effectiveness levels to identify the main variables determining governments' activeness to promote digital financial tools to address the current crisis.
3. Data and methods
3.1 Conceptual background and hypothesized model
In this paper, I perceive the government as a leader for change rather than a reactive agent. The earliest theory developed by Everett Rogers in 1962 “Diffusion of Innovation DOI” explains differences in adopting new technologies between innovators, early adopters, late adopters, early majority, late majority and laggards. This theory was also applied in the public policy, in which some countries are innovators of new technologies and other countries follow. However, this diffusion happens in regular and stable periods when there is no emergency.
Government response towards the COVID-19 crisis will have to go through three main phases (Eggers et al., 2020): respond, recover and thrive. For digital payments and online transactions to become the new normal in many developing countries, governments will have to go through these three phases. That being said, the evaluation of governments' response during this crisis is critical. This can enable us to understand better the future of digital payments over the next centuries. However, there is no guarantee that governments will turn their initial response into long-term plans.
In 1984, John Kingdon introduced the policy streams approach in which there are three main streams: the problem, the policy and the politics, which generate a cycle. In his theory, Kingdon and Stano (1984) see the problem or the crisis as an open “policy window” in which policymakers are somehow forced to take positive actions to solve other pending issues. An emerging problem becomes the centre of attention, depending on how rapidly it occurred, how intense it became and how much resources are needed to alter the situation (Gerston, 2014).
In this paper's context, governments are expected to use resources they already have (people, the government and businesses are already digitally included, and banks are already ready to serve). Thus, resources can be used to promote financial inclusion and digital payments.
Schumpeter's traditional theory describes the linkage between innovation and the adoption of innovative solutions and competition in finance (Witt, 2016). In his theory, for banks to generate profits and attract more customers and increase their market shares against other banks, banks will have to find innovative services and solutions to survive and achieve market dominance, which has also been confirmed in the study of Chandler (1990) . In Schumpeterian terms, any firm that survives in a competitive market faces uncertainty and instability and will have to innovate through applying new methods or use its resources differently. This can also be applied in the situation of a pandemic. In line with governments and central banks responses to the need for digital payments, commercial banks will also move towards applying modern innovative technologies (see Figure 1).
Therefore, this paper answers the following question: what determines governments' response to the surged need for digital payments resulted from the pandemic? To answer this question, I evaluate government responses to conclude whether the pandemic has encouraged low-income countries to adopt such policies or not. In this paper, I define five main dimensions which might explain countries’ propensity to take rapid actions related to digital transactions during the pandemic. The dimensions can be explained as follows: (1) financial institutions supply-side availability, variation and competition measured by the number of commercial banks in a country. (2) Government quality of policy formulation and implementation measured by government effectiveness. (3) Overall digital adoption of people, businesses and governments of all digitalized and electronic means of technology. (4) Face-to-face (physical) business transactions limitations explained by the closure of businesses and stores due to full lockdowns, which surged a need for electronic deals. (5) Income category of countries, which reflects country level of wealth and willingness to spend using new payment technologies. To empirically test the previous arguments, I set three main hypotheses:
The presence of policy response regarding digital payments during the pandemic depends on the economy's previous status of commercial banks, digital adoption and government effectiveness.
The presence of policy response regarding digital payments during the pandemic is higher in high-income countries.
Countries that experienced full lockdowns were more active in promoting digital payments than countries that did not.
3.2 Data definitions and sources
Table 1 summarizes the definitions of both dummy and continuous variables of this study.
The dependent variable of the study is G_Response which represents government response towards promoting digital means of payments. The supply side is represented through Num_Commercial_Banks variable, the effectiveness of governance is accounted for by including Governemnt_Effectiveness variable, the digital adoption is accounted for by including the Digital_Adoption variable and the variable Full_Lockdown represents the physical limitations imposed due to full-lockdowns during the pandemic.
In this study, I use data of 151 countries included in the financial related policy response dataset. This is a cross-sectional study in which data have been collected in 2020 after the global pandemic outbreak. Thus, my analysis has no regard to differences in time. Using this cross-sectional data, I compare the differences among different countries from different continents and different income levels with no unified policy response regarding digital payments.
The dependent variable is a dummy variable, which is given a value of (1) if the country had responded at least once to the pandemic by easing or promoting digital means of payments and (0) if not. I rely on five main variables to define the countries' characteristics, which showed a response regarding digital financial transactions during the pandemic (see Table 2).
In Table 3, descriptive statistics show that around 40% of the sample responded with policies to promote digital payments. Data show that around 60% of my sample belongs to upper-middle-income and high-income countries. Approximately 23% of the countries went in full lockdowns. The average country in this sample has a digital adoption score of 0.51 and an average of 38 commercial banks. The highest number of commercial banks (4,686) is in the United States, which was detected as an outlier. On average, in this sample, scores of government effectiveness is negative. Initial descriptive statistics show that 37% of the total countries considered upper-middle-income or high-income countries have responded, and 42% of the total countries are considered low-income lower-middle-income countries have responded. As expected, digital adoption and government effectiveness scores are both lower among low-income countries. Besides, more countries in the high-income group were under full-lockdown restrictions than lower-income countries. The average number of commercial banks is noticeably higher in high-income countries than in low-income countries (see Table 4).
Results of Table 5 show that only 59 countries showed a response, 56 % of them were high-income countries and 92 countries showed no response in which 61% were high-income countries. On average, regardless of the income category, countries that showed a response had higher government effectiveness levels, digital adoption, number of banks and lockdowns. Table 6 reveals the correlations between the variables of this study. A negative correlation showed up between Governemnt_Effectiveness, Digital_Adoption, Full_Lockdown and Num_Commercial_Banks with the variable Low_Income_Category, which is expected as low-income countries have weaker ICT infrastructure and weaker banking systems and weaker governments. However, there is a positive correlation between Low_Income_Category and G_Response. A positive correlation showed up between all variables and the government policy response variable except for Full_Lockdown; the strongest correlation is the number of commercial banks followed by Government_Effectiveness. Correlations results show full lockdowns amid the COVID-19 pandemic was lower in lower-middle-income and low-income countries. Results show that the strongest correlations are (1) the positive correlation between digital adoption and government effectiveness (0.70) and (2) the negative correlation between digital adoption and being a lower-middle-income or a low-income country (0.68). In discrete values, a simple cross-tabulation shows that 26 low- and lower-middle income countries actively responded [1] compared to 36, which did not, while 54 compared to 33 upper-middle-income or high-income countries showed a response, as seen in Table 7. On a categorical level, it is interesting to report that low-income countries (below $1036) and high-income countries (above $12535) are the only two categories that had more countries responding to the pandemic than not (see Table 7). The opposite showed up for lower- and upper-middle-income countries.
Policy responses varied among higher- and lower-income categories. Upper-middle and high-income countries policies were mainly focused on ensuring digital operational resilience and digital transactions security in the United States; Czech Republic; Estonia; France; Germany; Hungary; Italy; Latvia; Lithuania; Panama; Poland; Romania; Slovak Republic; Slovenia; Spain; Thailand; Slovak Republic; Bulgaria and Croatia. While utilizing know your customer (KYC) procedures and online documentation, remote bank, e-wallet account opening, remote financial services renewals, remote financial services documentation management or notarization were frequently used in upper-middle and high-income countries, such as in Argentina; Brazil; Maldives; Saudi Arabia; Jordan; Singapore; Thailand; Trinidad and Tobago; United States; Malaysia and Fiji, than in lower-income countries, such as Bangladesh; Nepal; Philippine; Afghanistan; and Sri Lanka.
Free distribution of credit cards/prepaid cards was a policy used by countries in the lower-income category only, such as Tunisia and Egypt. Using media or banks to promote for e-payments directly was another low-income and lower-middle income category special policy response, in countries such as Egypt; Bolivia; Congo; El Salvador; Senegal and Malawi in addition to limiting access to cash to promote digital payments, in countries such as Tunisia; Uganda and Egypt.
Increasing transaction limits was more frequently spotted in low-income and lower-middle income countries, such as Afghanistan; Indonesia; Liberia; Zambia Nepal and Ghana, than in upper-middle and high income countries, such as Botswana and Jordan only. The same goes for reducing transaction costs or waiving fees or suspending charges in which this policy was highly used in lower-middle income countries; such as Togo; Guinea-Bissau; Mali; Niger; Liberia; Uganda; Cote devoir; Zambia; Pakistan; Nepal; Ghana and Philippines, than in higher-income countries, such as Botswana; Russia and Indonesia.
Some policies were less frequently used in both higher- and lower-income groups; regulating unregulated digital and credit-only lenders or providing better online loans regulations was a less frequent policy response in the two-income categories, such asKenya and China. Using more point of sale (POS) terminals or more ATMs was another less frequent policy response in countries falling in the two-income categories, such as Egypt and Fiji. In addition to encouraging QR payments, in India, Egypt and Indonesia, transferring money through digital channels to vulnerable households to leverage digital payments infrastructure has been encouraged in Togo, Indonesia and Colombia.
3.3 Model specification
To identify key determinants of the government response during the pandemic, I first computed a binary variable indicating whether a country has taken a response or not. That is,
For my binary endogenous dependent variable, binary logistic regression model can be used. I first use a logistic link function as the following:
Using the same method in Equations (1) and (2), I test the following relationship:
In Equations (3) and (4), G_Response i is the dependent variable Y illustrated in Eqautions (1) and (2) in which (i = 1, …, 151). Since the dependent variable is a short-term immediate outstanding response rather than a long term attitude, there is no reason to believe that it can alter any of my independent variables of this study. Thus, endogeneity is less of a concern. Variables are expected to be positively associated with countries' decision regarding digital transactions during the pandemic except for the Low_Income_Category variable. In Equation (4), I use GDP per capita as a continuous variable instead of the Low_Income_Category binary variable to test my results' robustness. I expect GDP_percapita to be positively associated with government responses.
4. Results
Results of estimating variance inflation factors (VIFs) of my independent variables showed no evidence of multicollinearity, as seen in Table 8, which reveals no VIF of a value greater than 3.6. A tolerance less than 0.20 or a VIF of more than 10 is a concern (Menard, 1995; Hair et al., 1995). Sometimes a VIF of 10 is accepted and higher than 10 is a cause of concern (Marquardt, 1970; Mason et al., 1989). In more recent publications (Menard, 2001), a more conservative threshold is followed where VIF of more than 5 or sometimes more than 2.5 is considered problematic. I use the rule of 5 as a threshold. For my model, I used Pearson χ2 goodness-of-fit test to test H0 that there is no significant difference between the observed and the expected value against H1 that there is a significant difference between the observed and the expected value. The fitted model shows that my model fits reasonably well; (Prob > χ2 = 0.39 and 0.39). I used the Hosmer–Lemeshow test to test H0 that the observed and expected proportions are the same across all countries against H1 that they are not the same. Results show that my models fits well (Prob > χ2 = 0.49 and 0.71).
The regression was conducted with a White robust variance estimator. As seen in Table 9, the level of government effectiveness, falling in the low-income or lower-middle-income category and a number of commercial banks are three variables that should significantly impact government response to take actions towards digital transactions. In Table 10, replacing income category with GDP per capita showed the same results as higher GDP per capita has a negative impact. I estimated the marginal effects at means (see Table 11); in model (1), results show that the marginal effects of number of banks is very low (probability of making a policy response is only 0.2% with higher number of commercial banks); however, government effectiveness has a relatively higher magnitude (15% higher probability of taking actions regarding digital payments if a country has better government effectiveness levels). The impact of being in a low- or lower-income country is significantly higher; the probability that governments take actions related to facilitating financial digitalization during the pandemic is around 21% higher if the country is in a low- or lower- middle-income category. Figure 2 best describes the variations in marginal effects of all variables significant and insignificant variables.
In model (2), the results of Table 11 reveal that the probability of a response increases by 21% with higher government effectiveness levels. As for the GDP per capita, although having an extremely low impact on the response (<−0.000), the negative coefficient complies with the results of our base first model in which a categorical variable is more representative of country groupings than GDP per capita.
5. Discussion of the results
The results of this study has provided some evidence that governance responses to the COVID-19 crisis in low-income and lower-middle-income countries are moving these countries forward towards higher digital payments adoption levels. In total, 17 out of 26 countries that responded to the pandemic in low-income and lower-middle income categories are located in Africa. If this step has a long-term impact, one of the implications will be the contribution of the flourishing digital means of payments to these countries' economic growth, to the economic growth of Africa, and the greater openness they shall experience when engaged in the global market on both individual and business levels. Although low-income and lower-middle-income countries have weaker governing and banking systems, promoting digital payments seemed to be a vital option to maintain the level of consumption and collect dues to the government and to benefit from the contactless nature of digital payments. Governments found an opportunity to convert and financially monitor and include part of the formal digital one's huge informal cash economy. The abrupt acceptance of individuals of digital means of payments has also encouraged governments to offer more facilities to increase digital payments.
This study provided some evidence that the digital financial services in low- and lower middle-income countries are getting greater attention after the pandemic, which will entail new risks, including cybersecurity and low-quality credit booming. However, countries that will stay on track will benefit the most from this externality after the outside pressure caused by the pandemic eases. This will greatly depend on each country's long-term goals and the higher effort exerted to promote digital and financial transactions and the level of future investment in digital infrastructure, which will increase the level of financial inclusion. On the supply and legislative side, the lack of urgency had always led to delays in the movements towards digitizing payments; and the pandemic was shown as a catalyst for change. There is a shift in the behaviour and preferences on the demand side, which is one of the gains that will complement and support other efforts to achieve high digital literacy levels and digital financial literacy levels.
On the other hand, this sheds light on governments' impulsivity of low-income and lower-middle-income countries. Impulsive policy response during crisis and emergencies could signify a responsive government or a sign of lack of commitment to long-term plans, in which outside pressure is needed. It is yet to be studied whether the result of this study shows an attribute of “impulsivity” or “agility”. How impactful and progressive these countries will become will be revealed in the few upcoming years. According to the results of this study, regardless of the income category, in general, better governance practices and competitive banking systems are two predictors of a response during the COVID-19 pandemic to promote digital payments but digital adoption and full lockdowns are not.
6. Conclusion and policy suggestions
In this paper, I showed that the pandemic has led to a rise in the use of digital technologies as a result of the lockdown policies and restrictions on movements. The pandemic has impacted all aspects of life, and this digital surge has opened a discussion on the role of FinTechs, mobile money and all means of electronic financial transactions in containing some of the pandemic's negative outcomes. This study evaluates this topic from a public governance angle. The study finds that low- and lower-middle income countries had significantly responded during the pandemic compared to upper-middle-income and high-income countries. I have also found that in general, higher government effectiveness was a predictor of the response. The level of digital adoption of the countries before the pandemic showed no significant impact on their response neither during the pandemic nor the full lockdown.
Looking at the future, from a policy perspective, it is critical for countries, which already discerned the importance of digital payments in the modern economy, to immediately start drafting long-term digital resilience plans, to start considering digitization of money as a priority and to think of it as an innovative solution to address future economic disruptions. The lack of a bare minimum or basic and initial infrastructure required to even apply innovative solutions, such as mobile money wallets, indeed can hinder some countries from taking any actions in this regard. Thus, a country-specific, long-term plan to invest or attract foreign investment in digital and financial infrastructure is needed to foster a jumpstart in digitizing payments. The longer path for low-income and lower-middle-income countries will require them to embed digital and digital financial literacy in their culture through education and strengthening their formal banking systems. Besides, governments are encouraged to use the current crisis, which had produced an increased consumer appetite and a somehow forced familiarity with technology to transform to digitized government payments. If governments take the lead and become prime movers in this crisis, by setting a vision and easing and facilitating their regulations, many economies will experience a higher level of digital financial development in the future.
Figures
Variables definitions and sources
| Variable | Definition |
|---|---|
| G_Response | Response regarding digital transactions easing: (1) if at least one action was taken and (0) if no action was taken |
| Low_Income_Category | Country income level: (1) if a low- or lower-middle-income country level and (0) if an upper-middle or a high-income country level |
| Digital_Adoption | The worldwide index measures countries' digital adoption across three dimensions of the economy: people, government and business; a scale from 0 to 1 |
| Num_Commercial_Banks | Number of commercial banks in a country (Financial Access Survey) |
| Full_Lockdown | A dummy variable for businesses and stores closure due to full lockdowns during the pandemic: (1) if the country closed businesses and stores and (0) if not |
| Governemnt_Effectiveness | Estimates of units of a standard normal distribution; index ranges from −2.5 to 2.5 |
Note(s): *An international independent humanitarian analysis project; https://www.acaps.org/covid19-government-measures-dataset
Income classifications
Group | GNI per capita in current US$ | |
|---|---|---|
| 1 July 2020 (new) | 1 July 2019 | |
| Low income | <1,036 | <1,026 |
| Lower middle income | 1,036–4,045 | 1,026–3,995 |
| Upper middle income | 4,046–12,535 | 3,996–12,375 |
| High income | >12,535 | >12,375 |
Descriptive statistics of the total sample
| Variable | Obs | Mean | Std. dev | Min | Max |
|---|---|---|---|---|---|
| G_Response | 151 | 0.391 | 0.49 | 0 | 1 |
| Governemnt_Effectiveness | 147 | –0.131 | 0.84 | −1.909 | 2.231 |
| Low_Income_Category | 151 | 0.417 | 0.495 | 0 | 1 |
| Num_Commercial_Banks | 138 | 38.391 | 61.451 | 2 | 440 |
| Digital_Adoption | 140 | 0.508 | 0.177 | 0.147 | 0.871 |
| Full_Lockdown | 151 | 0.232 | 0.423 | 0 | 1 |
Descriptive statistics of high- and low-income countries
| Variable | Obs | Mean | Std. dev | Min | Max |
|---|---|---|---|---|---|
| Low/lower-middle income countries | |||||
| G_Response | 63 | 0.413 | 0.496 | 0 | 1 |
| Governemnt_Effectiveness | 62 | −0.659 | 0.59 | −1.909 | 1.342 |
| Num_Commercial_Banks | 54 | 23.981 | 24.108 | 2 | 161 |
| Digital_Adoption | 59 | 0.362 | 0.113 | 0.147 | 0.605 |
| Full_Lockdown | 63 | 0.175 | 0.383 | 0 | 1 |
| Higher-middle/high-income countries | |||||
| G_Response | 88 | 0.375 | 0.487 | 0 | 1 |
| Governemnt_Effectiveness | 85 | 0.253 | 0.786 | −1.847 | 2.231 |
| Num_Commercial_Banks | 84 | 47.655 | 75.099 | 2 | 440 |
| Digital_Adoption | 81 | 0.614 | 0.135 | 0.185 | 0.871 |
| Full_Lockdown | 88 | 0.273 | 0.448 | 0 | 1 |
Descriptive statistics of responding and non-responding countries
| Variable | Obs | Mean | Std. dev | Min | Max |
|---|---|---|---|---|---|
| Response | |||||
| Governemnt_Effectiveness | 59 | 0.095 | 0.866 | −1.555 | 2.231 |
| Low_Income_Category | 59 | 0.441 | 0.501 | 0 | 1 |
| Num_Commercial_Banks | 54 | 56.852 | 80.928 | 6 | 440 |
| Digital_Adoption | 56 | 0.532 | 0.187 | 0.16 | 0.871 |
| Full_Lockdown | 59 | 0.254 | 0.439 | 0 | 1 |
| No Response | |||||
| Governemnt_Effectiveness | 88 | −0.283 | 0.791 | −1.909 | 1.716 |
| Low_Income_Category | 92 | 0.402 | 0.493 | 0 | 1 |
| Num_Commercial_Banks | 84 | 26.524 | 41.068 | 2 | 321 |
| Digital_Adoption | 84 | 0.492 | 0.169 | 0.147 | 0.835 |
| Full_Lockdown | 92 | 0.217 | 0.415 | 0 | 1 |
Matrix of correlations between variables
| Variables | (1) | (2) | (3) | (4) | (5) | (6) |
|---|---|---|---|---|---|---|
| (1) G_Response | 1.000 | |||||
| (2) Gvt_Effectiveness | 0.169 | 1.000 | ||||
| (3) Low_Income_Category | 0.055 | −0.516 | 1.000 | |||
| (4) Num_Com_Banks | 0.234 | 0.311 | −0.197 | 1.000 | ||
| (5) Digital_Adoption | 0.090 | 0.706 | −0.681 | 0.409 | 1.000 | |
| (6) Full_Lockdown | −0.015 | −0.030 | −0.106 | −0.053 | 0.073 | 1.000 |
Cross-tabulation of countries over government responses
| Income level | Response: No | Response: Yes | Grand total |
|---|---|---|---|
| Low | 10 | 12 | 22 |
| Lower-middle | 26 | 14 | 40 |
| = 36 | = 26 | = 62 | |
| Upper-middle | 40 | 14 | 54 |
| High | 16 | 19 | 35 |
| = 56 | = 33 | = 89 | |
| Grand total | 92 | 59 | 151 |
Variance inflation factor
| VIF | 1/VIF | |
|---|---|---|
| Model (1) | ||
| Digital_Adoption | 3.049 | 0.328 |
| Government_Effectiveness | 2.037 | 0.491 |
| Low_Income_Category | 1.914 | 0.522 |
| Num_Commercial_Banks | 1.229 | 0.814 |
| Full_Lockdown | 1.033 | 0.968 |
| Mean VIF | 1.852 | |
| Model (2) | ||
| Digital_Adoption | 2.509 | 0.399 |
| GDP_percapita | 2.472 | 0.405 |
| Government_Effectiveness | 2.438 | 0.41 |
| Num_Commercial_Banks | 1.276 | 0.784 |
| Full_Lockdown | 1.031 | 0.97 |
| Mean VIF | 1.945 | |
Logit model results: Low_Income_Category as a predictor
| (1) | (2) | (3) | (4) | (5) | |
|---|---|---|---|---|---|
| Independent variables | |||||
| Government_Effectiveness | 0.557*** (0.213) | 0.898*** (0.255) | 0.633** (0.287) | 0.617* (0.346) | 0.628* (0.351) |
| Low_Income_Category | 0.981** (0.413) | 0.940** (0.412) | 0.887* (0.495) | 0.901* (0.496) | |
| Num_Commercial_Banks | 0.00794* (0.00454) | 0.00806* (0.00465) | 0.00812* (0.00460) | ||
| Digital_Adoption | −0.322 (1.849) | −0.366 (1.861) | |||
| Full_Lockdown | 0.127 (0.423) | ||||
| Constant | −0.344** (0.174) | −0.729*** (0.241) | −1.034*** (0.285) | −0.828 (1.080) | −0.844 (1.075) |
| Observations | 147 | 147 | 134 | 128 | 128 |
Note(s): ∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01
Logit model results: GDP_percapita as a predictor
| (6) | (7) | (8) | (9) | (10) | |
|---|---|---|---|---|---|
| Independent variables | |||||
| Government_Effectiveness | 0.557*** (0.213) | 0.841*** (0.302) | 0.912** (0.360) | 0.900** (0.393) | 0.901** (0.396) |
| GDP_percapita | −2.24e-05 (1.67e-05) | −5.74e-05** (2.29e-05) | −4.82e-05** (2.39e-05) | −4.82e-05** (2.38e-05) | |
| Num_Commercial_Banks | 0.0119** (0.00525) | 0.0117** (0.00529) | 0.0117** (0.00528) | ||
| Digital_Adoption | −0.936 (1.730) | −0.942(1.746) | |||
| Full_Lockdown | 0.0104 (0.411) | ||||
| Constant | −0.344** (0.174) | −0.0752 (0.269) | −0.181 (0.315) | 0.237 (0.891) | 0.237 (0.892) |
| Observations | 147 | 147 | 134 | 128 | 128 |
Note(s): ∗∗p < 0.05; ∗∗∗p < 0.01
Logit model marginal effect results
| dy/dx | Std. err | Z | P > z | [95% conf. interval] | ||
|---|---|---|---|---|---|---|
| Model (1) | ||||||
| Government_Effectiveness | 0.151 | 0.084 | 1.790 | 0.073 | −0.014 | 0.316 |
| Low_Income_Category | 0.217 | 0.119 | 1.820 | 0.069 | −0.017 | 0.451 |
| Num_Commercial_Banks | 0.002 | 0.001 | 1.740 | 0.082 | −0.000 | 0.004 |
| Digital_Adoption | −0.088 | 0.448 | −0.200 | 0.844 | −0.965 | 0.789 |
| Full_Lockdown | 0.031 | 0.102 | 0.300 | 0.764 | −0.169 | 0.230 |
| Model (2) | ||||||
| Government_Effectiveness | 0.217 | 0.095 | 2.290 | 0.022 | 0.031 | 0.403 |
| GDP_percapita | −0.000 | 0.000 | −2.030 | 0.043 | −0.000 | −0.000 |
| Num_Commercial_Banks | 0.003 | 0.001 | 2.180 | 0.029 | 0.000 | 0.005 |
| Digital_Adoption | −0.227 | 0.421 | −0.540 | 0.590 | −1.051 | 0.597 |
| Full_Lockdown | 0.003 | 0.099 | 0.030 | 0.980 | −0.192 | 0.197 |
Note
Countries are Liberia; Malawi; Burkina Faso; Uganda; Zambia; Côte d'Ivoire; Senegal; Kenya; Pakistan; Nepal; El Salvador; Ghana; Philippines; Bangladesh; Egypt, Arab Rep.; Indonesia; India; Afghanistan; Benin; Bolivia; Congo, Dem. Rep.; Guinea-Bissau; Mali; Niger; Togo; and Tunisia.
References
Aji, H.M., Berakon, I. and Husin, M.M. (2020), “COVID-19 and e-wallet usage intention: a multigroup analysis between Indonesia and Malaysia”, Cogent Business and Management, Vol. 7 No. 1, 1804181.
Allam, Z. (2020), “The forceful re-evaluation of cash-based transactions by COVID-19 and its opportunities to transition to cashless systems in digital urban networks”, Surveying the Covid-19 Pandemic and its Implications, pp. 107-117.
Asongu, S.A. and Biekpe, N. (2017), “Government quality determinants of ICT adoption in sub-Saharan Africa”, Netnomics: Economic Research and Electronic Networking, Vol. 18 Nos 2-3, pp. 107-130.
Auer, R., Cornelli, G. and Frost, J. (2020a), “COVID-19, cash, and the future of payments (No. 3)”, Bank for International Settlements.
Auer, R., Cornelli, G. and Frost, J. (2020b), Rise of the Central Bank Digital Currencies: Drivers, Approaches and Technologies, Monetary and Economic Department, Bank for International Settlements, Basel.
Augsburg, C. and Hedman, J. (2014), “Value added services and adoption of mobile payments”, ICEC '14: Proceedings of the Sixteenth International Conference on Electronic Commerce, pp. 27-32.
Ayo, C.K. and Babajide, D.O. (2006), “Designing a reliable E-payment system: Nigeria a case study”, Journal of Internet Banking and Commerce, Vol. 11 No. 2, pp. 1-3.
Bounie, D., Camara, Y. and Galbraith, J. (2020), “Consumers' mobility, expenditure and online-offline substitution response to COVID-19: evidence from French transaction data (No. 14-2020)”.
Capgemini (2019), “World payments report”, 15th ed., U.S. Department of Health and Human Services, available at: https://worldpaymentsreport.com/resources/world-payments-report-2019/.
Chandler, A. (1990), Scale and Scope: The Dynamics of Industrial Capitalism, The Belknap, Cambridge, Mass.
Chen, H., Engert, W., Huynh, K., Nicholls, G., Nicholson, M. and Zhu, J. (2020), “Cash and COVID-19: the impact of the pandemic on demand for and use of cash (No. 2020-6)”, Bank of Canada.
de Mello, L. and Ter-Minassian, T. (2020), “Digitalisation challenges and opportunities for subnational governments (No. 31)”, OECD Publishing.
Dubey, V., Sonar, R. and Mohanty, A. (2020), “FinTech, RegTech and contactless payments through the lens of COVID 19 times”, Money, Vol. 29 No. 6, pp. 3727-3734.
Eggers, W., Flynn, M., O'Leary, J. and Chew, B. (2020), “Government's response to COVID-19: from pandemic crisis to a better future”, A report from the Deloitte Centre for Government InsightsDeloitte Insights, New York.
Ellison, A., Williams, S. and Whyley, C. (2012), “The electronic payment needs of people on low incomes”, Payments Council Report.
Frost, J. (2020), The Economic Forces Driving FinTech Adoption across Countries, Netherlands Central Bank, Research Department, Basel.
Gerston, L.N. (2014), Public Policy Making: Process and Principles, Routledge; LSE.
Gharsellaoui, M. (2015), “The extent of competition and its impact on bank efficiency: case of the Tunisian commercial banks”, International Journal of Empirical Finance, Vol. 4 No. 5, pp. 278-290.
Hair, J.F. Jr, Anderson, R.E., Tatham, R.L. and Black, W.C. (1995), Multivariate Data Analysis, 3rd ed., Macmillan, New York.
Kingdon, J.W. and Stano, E. (1984), Agendas, Alternatives, and Public Policies, Vol. 45, Little, Brown, Boston, pp. 165-169.
Ligon, E., Malick, B., Sheth, K. and Trachtman, C. (2019), “What explains low adoption of digital payment technologies? Evidence from small-scale merchants in Jaipur, India”, PloS One, Vol. 14 No. 7, pp. 1-22.
Marquardt, D.W. (1970), “Generalized inverses, ridge regression, biased linear estimation, and nonlinear estimation”, Technometrics, Vol. 12, pp. 256-591.
Mason, R.L., Gunst, R.F. and Hess, J.L. (1989), Statistical Design and Analysis of Experiments: Applications to Engineering and Science, Wiley, New York.
Menard, S. (1995), “Applied logistic regression analysis (Sage University Paper Series on Quantitative Application in the Social Sciences, Series No. 106)”, 2nd ed., Sage, ThousandOaks, CA.
Menard, S. (2001), Applied Logistic Regression Analysis, 2nd ed., Sage Publications, London.
Menat, R. (2016), “Why we're so excited about FinTech”, The FinTech book: The Financial Technology Handbook for Investors, Entrepreneurs and Visionaries, pp. 10-12.
Moşteanu, D., Roxana, N., Faccia, D., Cavaliere, L.P.L. and Bhatia, S. (2020), “Digital technologies' implementation within financial and banking system during socio distancing restrictions–back to the future”, International Journal of Advanced Research in Engineering and Technology, Vol. 11 No. 6.
Patil, P.P., Dwivedi, Y.K. and Rana, N.P. (2017), “November). Digital payments adoption: an analysis of literature”, in Conference on E-Business, E-Services and E-Society, Springer, Cham, pp. 61-70.
Remolina, N. (2020), “Towards a data-driven financial system: the impact of COVID-19”, SMU Centre for AI & Data Governance, Research Paper, (08).
Schiliro, D. (2020), Towards Digital Globalization and the Covid-19 Challenge, University Library of Munich.
Sha'ban, M., Girardone, C. and Sarkisyan, A. (2020), “Cross-country variation in financial inclusion: a global perspective”, The European Journal of Finance, Vol. 26 Nos 4-5, pp. 319-340.
Singh, G. (2019), “A review of factors affecting digital payments and adoption behaviour for mobile e wallets”, International Journal of Research in Medical and Basic Sciences, Vol. 4 No. 6, pp. 89-96.
Strusani, D. and Houngbonon, G.V. (2020), “What COVID-19 means for digital infrastructure in emerging markets (No. 34306)”, The World Bank.
Thom, P.T. and Thuy, T.T.T. (2019), “Competition and efficiency among Vietnam's commercial banks”, Journal of Economic Development, Vol. 24 No. 1, pp. 104-118.
Tut, D. (2020), “FinTech and the COVID-19 pandemic: evidence from electronic payment systems (No. 102401)”, University Library of Munich.
Witt, U. (2016), “How evolutionary is Schumpeter's theory of economic development?”, in Rethinking Economic Evolution, Edward Elgar Publishing.
Further reading
Baumgartner, F.R. and Jones, B.D. (2010), Agendas and Instability in American Politics, University of Chicago Press.
Fu, J. and Mishra, M. (2020), “The global impact of COVID-19 on FinTech adoption (No. 20-38)”, Swiss Finance Institute.
Sweeny, K. (2008), “Crisis decision theory: decisions in the face of negative events”, Psychological Bulletin, Vol. 134 No. 1, p. 61.