Satisfaction with the COVID-19 Economic Stimulus Policy: A Study of the Special Cash Payment Policy for Residents of Japan

Abstract

The COVID-19 pandemic has caused a recession in the global economy. Many households have been experiencing financial difficulties. In order to curb its ripple effects on households, the Japanese government distributed a one-time cash payment of JPY 100,000 to every registered resident at the beginning of the pandemic. We utilized Hiroshima University’s Household Behavioral and Financial Survey in 2020 and 2021, and found that only 20% of the observations were satisfied with the one-time payment they received. We performed probit regression and found that those who were male, older, had a higher education level, and had anxiety about the future were less likely to be satisfied with this policy. Having a spouse, being a public official, and having higher household assets or better subjective health enhanced satisfaction with the policy. Our subsample analyses showed heterogeneity in satisfaction among various socioeconomic groups and in associations between their socioeconomic characteristics and policy satisfaction. The government should design future policies with due consideration of the challenges faced by various socioeconomic groups.

1. Introduction

The contagious nature of the new coronavirus has significantly impacted household finances worldwide [1,2,3,4,5,6,7]. As of 17 January 2022, the virus has infected over 326 million people and taken more than 5.5 million lives worldwide [8]. COVID-19 has undone decades of economic progress [9] and has worsened the prevailing socioeconomic challenges in most countries [10,11,12,13,14,15]. In Japan, the spread of COVID-19 led to the introduction of a nationwide state of emergency in April 2020 [16,17], a travel ban for foreign travelers [18,19], and the postponement of the 2020 Summer Olympics in Tokyo [20]. These changes pushed the Japanese economy into recession [21]. The Japanese government implemented a “special cash payment” policy to revive the economy. The government provided equal amounts of JPY 100,000 (equivalent to USD 872.79 as of 18 January 2022) to all registered resident as of 27 April 2020 [22,23]. The pandemic has affected individuals from different socioeconomic backgrounds in different ways [24,25,26]. One-time special cash payments may have been insufficient to help the most affected individuals, consequently caused lower policy satisfaction when compared with others. Therefore, the objective of this study was to explore the relationship between socioeconomic status and satisfaction with the special cash payment policy in Japan. The Japanese government is currently considering a similar payment policy [27]; therefore, exploring this relationship could help them identify the most vulnerable during the pandemic and execute more effective policies.

Several countries adopted nationwide fiscal stimulus payment policies during the pandemic such as Japan, the United States, South Korea, Singapore, and Hong Kong SAR [28]. Among them, the United States’ stimulus payments policy is comparable to the Japanese policy. According to the U.S. Department of the Treasury (USDT), the federal government provided three rounds of progressive direct subsidies nationwide [29]. At the beginning of the pandemic, the government provided up to USD 1200 to an eligible adult and up to USD 500 to an eligible child aged below 17 years [30]. Then, the government distributed a second payment in December 2020, which disbursed up to USD 600 to each eligible adult and each eligible child aged below 17 years [30]. Finally, a third payment was distributed in March 2021 and disbursed up to USD 1400 to each eligible adult [30]. The government provided additional payments of up to USD 1400 per dependent [30]. The Singaporean government provided several rounds of cash payments to adults aged 21 years and above [31]. Ultimately, an average Singaporean adult could receive a total payment of up to SGD 1200 in total [31], which is equivalent to USD 888.73 as of January 2022. In South Korea, the government provided a subsidy to households in the bottom 70% of the income bracket. Households with at least four members received direct payments worth KRW 1 million each (equivalent to USD 838.83 as of 18 January 2022) [32]. In contrast, smaller households received scaled-down payments [32]. Finally, the Spanish government decided to reduce the COVID-19 income shock with a universal basic income policy called the “Minimum Vital Income” [33]. This policy enabled an eligible adult to receive monthly payments worth one-twelfth of the annual sum of a non-contributory pension, equivalent to EURO 469.93 or USD 533.98 (as of January 2022) in 2021 [33].

Although stimulus payments have been widely implemented to alleviate the impact of the pandemic, similar policies have been implemented to mitigate the residuum of previous financial crises. During the Great Recession, the U.S. Federal Government sent USD 600 (USD 779.73 in 2022) to every single tax-filer and USD 1200 (USD 1559.46 in 2022) to each married joint tax-filer [34]. Sahm et al. [35] found that more than half of the respondents had used the stimulus payment for debt settlement and 23% of the respondents had increased their spending. It has been estimated that the increased spending may have been worth around USD 32 billion (USD 41.59 billion in 2022) [35]. In Taiwan, the government sought to reduce the impact of the Great Recession by distributing a shopping coupon worth NTD 3600 (equivalent to USD 107.37 as of January 2009 or USD 139.53 in 2022) to each person [36]. Although Lin and Chen [37] argued that the policy was unsuccessful in stimulating the Taiwanese economy, Kan et al. [38] found that the policy helped increase unplanned spending by 24%. In Japan, the government had also distributed shopping coupons in 1999 [39], 2009, and 2015 [40]. Several studies have identified the beneficial effect of the coupons [39,40]: Kadoya et al. [40] suggested that people of different socioeconomic backgrounds responded to the policy differently.

Studies have found that people of different socioeconomic statuses can experience the same effects at different magnitudes [41,42,43,44,45]. Some have also suggested that different people respond differently to cash transfers [43,46,47,48,49]. In the United States, lower-income households have been experiencing the COVID-19 income shock at a greater magnitude than higher-income households [46,49]. Studies by Chetty et al. [46] and Mahjabeen and Pratoomchat [48] found that higher-income households are more likely to save the cash subsidies received; in contrast, lower-income households are more likely to spend cash subsidies on daily expenses. Perez-Lopez and Monte [49] found that some households used their subsidies to settle debts they incurred as a result of COVID-19. COVID-19 income shocks have affected Japanese people with different socioeconomic backgrounds differently [43,50]. People with lower incomes are more likely to experience more severe economic shocks [50]. Those with low income are typically young, female, and less skilled [50], or have liquidity constraints [43]. The pandemic has relentlessly disrupted the service industries, with which these more vulnerable groups are associated; therefore, they are more likely to experience financial losses [50]. As a result, they are more likely to experience mental health problems and suicidal thoughts [51]. A one-time special cash payment policy in Japan may be insufficient to alleviate the effect of the income shock that these vulnerable people have been experiencing.

This study focused on the relationship between socioeconomic status and satisfaction with the special cash payment policy. Although the Japanese government and others have implemented other economic stimulus measures such as industry-level financial aid [30,52], SME loans [53,54], consumption coupons [55], direct business subsidies programs [55], job retention schemes [56], and travel subsidies programs [57], the impacts of these policies toward individuals are difficult to observe or detangle. The direct subsidy program has been implemented at the mass level, which makes it vital to observe satisfaction towards the government policy regarding the pandemic. Focusing on the relationship between socioeconomic status and satisfaction toward the unconditional cash transfer policy can help us identify the most vulnerable quickly and efficiently. Intrinsically, the cash subsidy may leave people feeling satisfied. However, the level of satisfaction will vary across socioeconomic groups. We hypothesize that female, young, and low-income people are more likely to have dissatisfaction with the one-time cash transfer policy. People with a higher level of economic vulnerability may find that a one-time direct payment of JPY 100,000 is insufficient to compensate for the economic loss they faced during and as a result of the pandemic. They are more likely to have lower levels of satisfaction with the policy than others. Although our measure of satisfaction is subjective, several studies, such as those by AlKhaldi et al. [58], DeHoog et al. [59], Friedberg et al. [60], Herman [61], Kampen et al. [62], Lazarus et al. [63], Oleribe et al. [64], and Ybema et al. [65], used similar methods. The main approaches to identifying vulnerable households are time-consuming; therefore, this method can help us identify the most vulnerable group in a timely manner during the pandemic in Japan and prevent them from suffering a harsher economic shock. To the best of our knowledge, this study is the first to identify those vulnerable to the economic shock of COVID-19 by studying the relationship between socioeconomic status and satisfaction with the direct household subsidy program in Japan. This study can guide policymakers in Japan in developing more comprehensive policies to alleviate the ongoing effects of the economic recession on vulnerable households in Japan. This study is comparable across countries because of the nature of the policy. In contrast, other policies are based on existing ones that vary from country to country. The effectiveness of the direct payment program can be compared internationally. With these benefits, this study provides essential insights for policymakers worldwide on how their counterparts elsewhere have strived to address economic hardship during the pandemic.

The rest of this paper is organized as follows. Section 2 describes the data and methods, Section 3 presents the empirical results, Section 4 discusses the results, and Section 5 concludes the paper.

2. Data and Methods

2.1. Data

This study used panel data from Hiroshima University’s Household Behavioral and Financial Survey, managed by the Hiroshima Institute of Health Economics Research (HiHER). The database comprises datasets from two waves of online surveys. The first round of data collection took place from 20 to 25 February 2020, which was at the beginning of the COVID-19 pandemic. The second round took place from 19 to 26 February 2021. Both rounds were conducted by Nikkei Research, one of the most extensive databases representing the Japanese population across socioeconomic status. Following the random sampling procedure, the survey collected information on the preferences and socioeconomic statuses of 17,463 Japanese adults who answered the survey in 2020 and the 6103 of them who answered the survey again in 2021. In both waves, the representativeness of the data was ensured through the proper representation of respondents from all demographic and socioeconomic strata.

Our study focuses on the relationship between satisfaction with the special cash payment policy and socioeconomic status. Therefore, we used the 2021 dataset, which comprised information on the satisfaction of the respondent with the policy, marital status, household size, career type, financial status, and perspectives on health and the future. We also used the 2020 dataset, which included basic demographic information such as gender, age, educational attainment, residential area, and child-rearing status. Due to missing data on interesting variables such as household financial status, we excluded 2331 observations. Our final sample comprised 3772 observations, which represented 61.81% of the observations for 2021.

2.2. Variable Definitions

Our dependent variable, “Satisfaction with special cash payment,” was binary. It was based on the survey question for 2021, which asked the respondent to rate the perception of the statement, “The special cash payment of 100,000 yen by the Government of Japan last year was sufficient for me to cope with my anxiety about household finance.” The respondent could rate it on a five-point scale: 1 = It does not hold true at all for you; 2 = It is not so true for you; and 3 = Neither; to 4 = It is rather true for you; and 5 = It is particularly true for you. However, we converted the ordinal response to a binary response to measure satisfaction and dissatisfaction with cash transfer policy more concretely. If the respondents answered 4 or 5, we classified them as “satisfied with the policy,” i.e., “Satisfaction toward special cash payment” = 1. Otherwise, we coded “Satisfaction towards special cash payment” = 0.

Our explanatory variables were mainly associated with the socioeconomic status, such as gender, age, location of residence, years of education, marital and childcare status, and household size, income, and asset value. We included subjective health status, and anxieties about and myopic views of the future.

Table 1. Variable definitions.

Variables	Definition
Dependent variable
Satisfaction with the special cash payment	Binary variable: 1 = It is rather true for you, or it is particularly true for you for the statement “The special cash payment of 100,000 yen from the Government of Japan last year was sufficient for me to cope with my anxiety around household finance.” and 0 = otherwise
Explanatory variables
Male *	Binary variable: 1 = Male and 0 = Female
Age *	Discrete variable: Respondent’s age
Age squared	Discrete variable: Respondent’s age squared
Highly populated city *	Binary variable: 1 = Living in special wards in Tokyo or government-designated city areas and 0 = otherwise
Years of education *	Discrete variable: Years of education
Spouse	Binary variable: 1 = Currently have a spouse or partner and 0 = otherwise
Having a child/children *	Binary variable: 1 = Having a child/children and 0 = otherwise
Household size	Discrete variable: Number of people currently living in the household
Working as a public official	Binary variable: 1 = Currently working as a public official and 0 = otherwise
Household income	Continuous variable: Annual earned income before taxes and with bonuses of the entire household in 2020 (unit: JP¥)
Log of household income	Log (household income)
Household assets	Continuous variable: Balance of financial assets (savings, stocks, bonds, insurance, etc.) of entire household (unit: JP¥)
Log of household assets	Log (household assets)
Subjective health	Ordinal variable: 1 = It does not hold true at all for you; 2 = It is not so true for you; 3 = Neither true nor not true; 4 = It is rather true for you; 5 = It is particularly true for you for the statement “I am now healthy and was generally healthy in the last one year.”
Future anxiety	Ordinal variable: 1 = It does not hold true at all for you; 2 = It is not so true for you; 3 = Neither true nor not true; 4 = It is rather true for you; 5 = It is particularly true for you for the statement “I have anxieties about ‘life after 65 years of age’ (For those who were already aged 65 years or above, ‘life in the future’).”
Myopic view of the future	Ordinal variable: 1 = Completely disagree; 2 = Disagree; 3 = Neither agree nor disagree; 4 = Agree; 5 = Completely agree with the statement “As the future is uncertain, it is a waste to think about it.”

Note: * Denotes the data from the 2020 wave.

defines these variables.

2.3. Descriptive Statistics

Our sample comprised 3772 observations. Descriptive statistics (

Table 2. Descriptive statistics.

Variable	Mean	Median	Std. Dev.	Min	Max
Dependent variables
Satisfaction with the special cash payment	0.20	0	0.40	0	1
Explanatory variables
Male	0.67	1	0.47	0	1
Age	50.96	51.00	13.63	21.00	86.00
Age squared	2782.87	2601.00	1404.70	441.00	7396.00
Highly populated city	0.42	0	0.49	0	1
Years of education	15.01	16	2.10	9	21
Having a child/children	0.58	1	0.49	0	1
Spouse	0.67	1	0.47	0	1
Household size	2.58	2	1.22	1	10
Working as a public official	0.06	0	0.23	0	1
Household income	6,458,378	5,000,000	4,134,713	500,000	21,000,000
Log of household income	15.45	15.42	0.75	13.12	16.86
Household assets value	21,000,000	8,750,000	29,800,000	1,250,000	125,000,000
Log of household assets	15.92	15.98	1.44	14.04	18.64
Subjective health status	3.26	3	1.08	1	5
Future anxiety	3.69	4	1.14	1	5
Myopic view of the future	2.67	3	1.02	1	5
Observations		3772

) revealed that 67% of the respondents were men, and 42% lived in designated cities with a population greater than 500,000 inhabitants. The average age was 51 years and the average period of education was 15 years. An average respondent lived in a household of 2.58 people with an annual income of JPY 6.45 million (around USD 56,301.41 as of January 2022) and assets of JPY 21 million (around USD 183,306.9 as of 18 January 2022). Most of the respondents (67%) had a spouse or partner, and 58% had children. Only 6% were public officials.

Table 2 shows that the respondents were, on average, dissatisfied with the special cash payment policy. Only 20% were satisfied.

Table 3. Distribution of satisfaction with the special cash payment policy by age group.

Satisfaction with the Special Cash Payment Policy	Age				Total
	<=35 Years	36–50 Years	51–65 Years	>65 Years
Satisfied	143	253	218	122	736
	24.28%	19.63%	17.26%	19.33%	19.51%
Otherwise	446	1036	1045	509	3036
	75.72%	80.37%	82.74%	80.67%	80.49%
Total	589	1289	1263	631	3772
	100%	100%	100%	100%	100%
Mean difference	F = 4.22 ***

Note: *** p < 0.01.

provides the distribution of satisfaction with the special cash payment policy by age group. Then, we set a null hypothesis where group means were equal, i.e., satisfaction toward the special cash payment policy was equal across all age groups. We used one-way ANOVA to test our hypothesis and calculated the F-value, which was equal to 4.22 (Table 3). The F-statistic (3, ∞) at a 0.01 significance level was 3.79; therefore, the null hypothesis was rejected. Specifically, there were significant differences in satisfaction with the policy by age group. A higher proportion of younger respondents was satisfied with the policy. Overall, 24% and 17% to 20% of the respondents aged 35 years and below, and in older cohorts, were satisfied with the policy, respectively. The cohort aged between 51 and 65 years had the lowest level of satisfaction.

Table 4. Distribution of satisfaction with the special cash payment policy by gender, city status, and current occupation as a public official.

Satisfaction with the Special Cash Payment Policy	Gender		City Status		Public Official		Total
	Male	Female	City	Rural	Yes	No
Satisfied	451	285	297	439	52	684	736
	17.88%	22.80%	18.69%	20.11%	24.53%	19.21%	19.51%
Otherwise	2071	965	1292	1744	160	2876	3036
	82.12%	77.20%	81.31%	79.89%	75.47%	80.79%	80.49%
Total	2522	1250	1589	2183	212	3560	3772
	100%	100%	100%	100%	100%	100%	100%
Mean difference	t = 3.5924 ***		t = 1.0857		t = −1.8974 *

Note: *** p < 0.01, * p < 0.10.

provides the distribution of satisfaction with the special cash payment policy by gender, city status, and current occupation as a public official. We performed t-tests, assuming equal means as the null hypothesis. We also calculated t-statistics; these values are shown in Table 4. Our t-tests revealed significant differences in satisfaction levels by gender at the 99% confidence level. Women (23%) were more satisfied with the policy than men (18%). Our t-tests also revealed significant differences by stable career status at the 90% confidence level. Public officials (25%) were more satisfied with the policy than others (19%). There were no statistical differences in satisfaction among respondents who lived in highly populated cities. Our t-tests revealed that the t-value of 1.0857 was smaller than the critical t-value of 1.646 at the 90% confidence level.

2.4. Methods

Although the descriptive statistics (Table 2) and distributions (Table 3 and Table 4) provide some details on the characteristics of the respondent in relation to policy satisfaction, we estimated the effects of each characteristic while keeping the others constant.

Our dependent variable, “Satisfaction with the special cash payment policy” was binary; therefore, we used probit regression to estimate the following equation:

$$Y_i=f\left(X_i,{\epsilon }_i\right)$$

where $Y_i$ is satisfaction with the special cash payment policy, $X$ is a vector of the respondent’s characteristics, and $\epsilon$ is the error term. The probit model helps to estimate the probability that an observation with specific features will fall into either of the outcome categories. Thus, we believe that the probit model will appropriately estimate the likelihood of respondents with specific socioeconomic backgrounds to feel satisfaction or dissatisfaction toward special cash payment policy. Previous studies which have used binary responses as dependent variables also used probit regression model [40].

Our model specifications are as follows:

• $$\begin{gathered} Satisfaction toward special cash payment= {\beta }_0+{\beta }_{1}mal{e}_i+{\beta }_{2}ag{e}_i+ \\ {\beta }_{3}age square{d}_i+{\beta }_{4}high population cit{y}_i+{\beta }_{5}years of educatio{n}_i+{\epsilon }_i; \end{gathered}$$
• $$\begin{gathered} Satisfaction toward special cash payment= {\beta }_0+{\beta }_{1}mal{e}_i+{\beta }_{2}ag{e}_i+ \\ {\beta }_{3}age square{d}_i+{\beta }_{4}high population cit{y}_i+{\beta }_{5}years of educatio{n}_i+ \\ {\beta }_{6}having childre{n}_i+{\beta }_{7}spous{e}_i+{\beta }_{8}household siz{e}_i+{\epsilon }_i; \end{gathered}$$
• $$\begin{gathered} Satisfaction toward special cash payment= {\beta }_0+{\beta }_{1}mal{e}_i+{\beta }_{2}ag{e}_i+ \\ {\beta }_{3}age square{d}_i+{\beta }_{4}high population cit{y}_i+{\beta }_{5}years of educatio{n}_i+ \\ {\beta }_{6}having childre{n}_i+{\beta }_{7}spous{e}_i+{\beta }_{8}household siz{e}_i+{\beta }_{9}public officia{l}_i+ \\ {\beta }_{10}\log (household incom{e}_i)+{\beta }_{11}\log (household asset{s}_i)+{\epsilon }_i; \end{gathered}$$
• $$\begin{gathered} Satisfaction toward special cash payment= {\beta }_0+{\beta }_{1}mal{e}_i+{\beta }_{2}ag{e}_i+ \\ {\beta }_{3}age square{d}_i+{\beta }_{4}high population cit{y}_i+{\beta }_{5}years of educatio{n}_i+ \\ {\beta }_{6}having childre{n}_i+{\beta }_{7}spous{e}_i+{\beta }_{8}household siz{e}_i+{\beta }_{9}public officia{l}_i+ \\ {\beta }_{10}\log (household incom{e}_i)+{\beta }_{11}\log (household asset{s}_i)+{\beta }_{12}subjective healt{h}_i+ \\ {\beta }_{13}future anxiet{y}_i+{\beta }_{14}myopic view of the futur{e}_i+{\epsilon }_i. \end{gathered}$$

The first model is our base. The explanatory variables comprise basic demographic variables such as gender, age, age squared, years of education, and living area. The second model introduces the household member variables, such as having a child/children, having a spouse, and household size. We included the employment and financial status variables in the third model, and subjective health and anxiety about and myopic view of the future in the fourth model.

The explanatory variables in the models could potentially be multicollinear; therefore, we conducted multicollinearity tests in all models (the results have not been reported here to save on space, but are available upon request). The values of the variance inflation factors of the explanatory variables were below 10, indicating that multicollinearity was not significant in all models.

3. Results

Section 3.1 presents the results of the full sample analysis. Section 3.2 and Section 3.3 provide subsample analysis by age and gender, respectively.

3.1. Full Sample Analysis

Table 5. Probit model regression results (full sample analysis).

Variables	Dependent Variable: Satisfaction with the Special Cash Payment Policy
	Model 1	Model 2	Model 3	Model 4
Male	−0.117 **	−0.111 **	−0.111 **	−0.108 **
	(0.0530)	(0.0533)	(0.0537)	(0.0544)
Age	−0.0407 ***	−0.0404 ***	−0.0438 ***	−0.0346 ***
	(0.0116)	(0.0117)	(0.0118)	(0.0121)
Age squared	0.000355 ***	0.000349 ***	0.000369 ***	0.000277 **
	(0.000113)	(0.000115)	(0.000116)	(0.000119)
High population city	−0.0439	−0.0424	−0.0475	−0.0650
	(0.0479)	(0.0480)	(0.0483)	(0.0489)
Years of education	−0.00635	−0.00899	−0.0216 *	−0.0244 **
	(0.0113)	(0.0114)	(0.0121)	(0.0123)
Having a child/children		−0.0774	−0.0734	−0.113
		(0.0678)	(0.0681)	(0.0691)
Spouse		0.161 **	0.144 **	0.145 **
		(0.0656)	(0.0674)	(0.0679)
Household size		−0.00816	−0.0157	−0.00324
		(0.0250)	(0.0253)	(0.0259)
Public official			0.173 *	0.175 *
			(0.0983)	(0.0993)
Log of household income			0.0272	−0.00630
			(0.0405)	(0.0408)
Log of household assets			0.0503 **	0.0372 *
			(0.0198)	(0.0207)
Subjective health				0.221 ***
				(0.0241)
Future anxiety				−0.0470 **
				(0.0226)
Myopic view of the future				0.0242
				(0.0244)
Constant	0.413	0.397	−0.504	−0.582
	(0.329)	(0.330)	(0.586)	(0.640)
Observations	3772	3772	3772	3772
Log-likelihood	−1846	−1843	−1836	−1783
Chi2 statistics	30.75	36.20	49.27	141.5
p-value	0.0000	0.0000	0.0000	0.0000

Note: *** p < 0.01, ** p < 0.05, * p < 0.10.

presents the estimation results of the model specifications (1) to (4). The probit regression results show that being male, age, years of education, and future anxiety had significantly negative associations, whereas age squared, having a spouse, being a public official, household assets, and subjective health status had significantly positive associations with satisfaction with the special cash payment policy. The significant age squared variable indicates that the association between special cash payment policy and age is not linear. The results are primarily robust and consistent in the model specifications and show that when one is male or older, they are more likely to be dissatisfied with the special cash payment policy. Education and future anxiety negatively influenced policy satisfaction. Having a spouse, working as a public official, household assets, and subjective health positively affected policy satisfaction.

We discussed in the data sub-section that we converted the ordinal response to a binary response when measuring the binary dependent variable. To check the robustness of our results, we also ran the ordered probit regression, considering the original ordinal responses as the dependent variable. The results of the ordered probit regression models are mostly similar to the probit regression model. Due to the similarity, we did not report the results of the ordered probit model to save on space, although these are available upon request.

3.2. Subsample Analysis by Gender

We analyzed the subsample by gender, and found that it had a significant effect on policy satisfaction.

Table 6. Probit model regression results (subsample analysis: gender).

Variables	Dependent Variable: Satisfaction with the Special Cash Payment
	Subsample: Male				Subsample: Female
	Model 1	Model 2	Model 3	Model 4	Model 1	Model 2	Model 3	Model 4
Age in 2021	−0.0445 ***	−0.0430 ***	−0.0458 ***	−0.0353 **	−0.0327	−0.0390 *	−0.0374 *	−0.0231
	(0.0157)	(0.0158)	(0.0159)	(0.0163)	(0.0200)	(0.0204)	(0.0203)	(0.0209)
Age square	0.000393 ***	0.000369 **	0.000387 **	0.000285 *	0.000269	0.000349	0.000318	0.000169
	(0.000147)	(0.000149)	(0.000150)	(0.000154)	(0.000210)	(0.000216)	(0.000216)	(0.000222)
High population city	−0.0402	−0.0456	−0.0555	−0.0682	−0.0484	−0.0267	−0.0231	−0.0474
	(0.0597)	(0.0599)	(0.0602)	(0.0609)	(0.0802)	(0.0808)	(0.0812)	(0.0822)
Years of education	−0.0108	−0.0125	−0.0263 *	−0.0278 *	0.00313	0.00179	−0.00744	−0.0158
	(0.0135)	(0.0136)	(0.0144)	(0.0146)	(0.0210)	(0.0210)	(0.0226)	(0.0226)
Having child		−0.0420	−0.0426	−0.0611		−0.131	−0.127	−0.199 *
		(0.0896)	(0.0902)	(0.0909)		(0.105)	(0.106)	(0.108)
Spouse		0.134	0.108	0.109		0.194 **	0.208 **	0.183 *
		(0.0900)	(0.0915)	(0.0926)		(0.0979)	(0.103)	(0.104)
Household size		−0.0317	−0.0430	−0.0328		0.0381	0.0356	0.0529
		(0.0319)	(0.0322)	(0.0329)		(0.0409)	(0.0416)	(0.0422)
Public official			0.0883	0.0943			0.487 **	0.475 **
			(0.115)	(0.116)			(0.201)	(0.202)
Log of household income			0.0561	0.0192			−0.0142	−0.0374
			(0.0517)	(0.0522)			(0.0656)	(0.0660)
Log of a household assets value			0.0531 **	0.0394			0.0346	0.0227
			(0.0239)	(0.0250)			(0.0362)	(0.0377)
Subjective health				0.211 ***				0.245 ***
				(0.0304)				(0.0398)
Future anxiety				−0.0605 **				−0.0227
				(0.0283)				(0.0379)
The myopic view of the future				0.0354				0.00319
				(0.0307)				(0.0407)
Constant	0.452	0.456	−0.929	−0.960	0.107	0.106	−0.121	−0.485
	(0.457)	(0.459)	(0.785)	(0.846)	(0.544)	(0.548)	(0.936)	(1.037)
Observations	2522	2522	2522	2522	1250	1250	1250	1250
Log-likelihood	−1179	−1178	−1173	−1140	−666.8	−663.7	−660.2	−639.3
Chi2 statistics	10.07	12.44	22.91	81.01	8.464	13.86	20.92	58.01
p-value	0.0393	0.0871	0.0111	0.0000	0.0760	0.0537	0.0217	0.0000

Note: *** p < 0.01, ** p < 0.05, * p < 0.10.

presents the results. The probit regression results for males show that age, years of education, and future anxiety had significantly negative associations, whereas age squared, household assets, and subjective health status had significantly positive associations with satisfaction with the special cash payment policy. For females, age and having a child had significantly negative associations, whereas having a spouse, being a public official, and subjective health status had significantly positive associations with satisfaction with the special cash payment policy. The effects of age and age squared were consistent and robust among male respondents. Future anxiety affected only male respondents’ satisfaction with the special cash payment policy. The effects of having a spouse and being a public official were consistent and robust among female respondents. The effect of subjective health alone was observable in both genders.

3.3. Subsample Analysis by Age Group

Age significantly affected policy satisfaction; therefore, we performed subsample analysis by age group.

Table 7. Probit model regression results (subsample analysis: Age group 35–50).

Variables	Dependent Variable: Satisfaction with the Special Cash Payment Policy
	Subsample: Age <= 35 Years				Subsample: 35 Years < Age 50 Years
	Model 1	Model 2	Model 3	Model 4	Model 1	Model 2	Model 3	Model 4
Male	−0.179	−0.142	−0.178	−0.221 *	−0.102	−0.103	−0.115	−0.110
	(0.122)	(0.125)	(0.125)	(0.129)	(0.0841)	(0.0844)	(0.0850)	(0.0873)
Age in 2021	0.122	−0.0456	−0.0255	−0.0206	0.176	0.192	0.227	0.187
	(0.255)	(0.261)	(0.258)	(0.266)	(0.210)	(0.210)	(0.212)	(0.217)
Age squared	−0.00283	−6.51e-05	−0.000523	−0.000469	−0.00237	−0.00257	−0.00299	−0.00250
	(0.00434)	(0.00442)	(0.00437)	(0.00450)	(0.00242)	(0.00243)	(0.00245)	(0.00251)
Highly populated city	−0.0703	−0.0442	−0.0395	−0.0433	−0.0100	−0.0166	−0.00405	−0.0403
	(0.117)	(0.120)	(0.121)	(0.123)	(0.0824)	(0.0829)	(0.0836)	(0.0849)
Years of education	0.0127	0.00357	−0.00851	−0.0112	−0.0185	−0.0177	−0.0310	−0.0395 **
	(0.0290)	(0.0303)	(0.0318)	(0.0320)	(0.0182)	(0.0182)	(0.0195)	(0.0196)
Having a child/children		−0.226	−0.207	−0.263		0.116	0.104	0.0696
		(0.159)	(0.161)	(0.166)		(0.120)	(0.121)	(0.124)
Spouse		0.470 ***	0.511 ***	0.495 ***		0.00565	−0.0143	−0.0291
		(0.137)	(0.147)	(0.151)		(0.115)	(0.118)	(0.120)
Household size		−0.0503	−0.0524	−0.0377		−0.0456	−0.0489	−0.0429
		(0.0575)	(0.0597)	(0.0597)		(0.0397)	(0.0399)	(0.0411)
Working as a public official			0.548 ***	0.561 ***			0.306 *	0.245
			(0.201)	(0.205)			(0.171)	(0.171)
Log of household income			−0.119	−0.166			0.0503	0.0329
			(0.102)	(0.105)			(0.0740)	(0.0754)
Log of household assets			0.131 **	0.124 **			0.0261	0.0226
			(0.0609)	(0.0628)			(0.0345)	(0.0355)
Subjective health				0.195 ***				0.272 ***
				(0.0612)				(0.0416)
Future anxiety				−0.0587				0.0151
				(0.0564)				(0.0400)
Myopic view of the future				0.0322				0.0345
				(0.0581)				(0.0422)
Constant	−1.877	0.625	0.386	0.561	−3.643	−3.932	−5.651	−5.437
	(3.690)	(3.759)	(3.917)	(3.971)	(4.511)	(4.524)	(4.659)	(4.775)
Observations	589	589	589	589	1289	1289	1289	1289
Log-likelihood	−321.1	−315.6	−309.5	−303.2	−631.7	−630.9	−628.2	−602.5
Chi2 statistics	11.13	23.87	35.34	47.14	13.73	15.73	21.24	63.06
p-value	0.0489	0.0024	0.0002	0.0000	0.0174	0.0464	0.0309	0.0000

Note: *** p < 0.01, ** p < 0.05, * p < 0.10.

and

Table 8. Probit model regression results (subsample analysis: Age group 50–65).

Variables	Dependent Variable: Satisfaction with the Special Cash Payment Policy
	Subsample: 50 Years < Age 65 Years				Subsample: Age > 65 Years
	Model 1	Model 2	Model 3	Model 4	Model 1	Model 2	Model 3	Model 4
Male	−0.141	−0.140	−0.106	−0.106	−0.00936	−0.0312	−0.0236	0.0262
	(0.0977)	(0.0987)	(0.100)	(0.103)	(0.161)	(0.165)	(0.167)	(0.170)
Age in 2021	0.313	0.324	0.245	0.291	−0.222	−0.220	−0.230	−0.122
	(0.294)	(0.295)	(0.299)	(0.308)	(0.364)	(0.364)	(0.365)	(0.371)
Age squared	−0.00289	−0.00298	−0.00234	−0.00274	0.00163	0.00161	0.00168	0.000915
	(0.00255)	(0.00256)	(0.00259)	(0.00267)	(0.00248)	(0.00248)	(0.00249)	(0.00253)
Highly populated city	−0.0117	−0.00913	−0.0316	−0.0491	−0.147	−0.148	−0.151	−0.170
	(0.0843)	(0.0846)	(0.0851)	(0.0863)	(0.120)	(0.121)	(0.121)	(0.122)
Years of education	−0.00272	−0.00418	−0.0191	−0.0179	−0.00524	−0.00720	−0.00875	−0.00857
	(0.0211)	(0.0212)	(0.0227)	(0.0233)	(0.0281)	(0.0283)	(0.0290)	(0.0299)
Having a child/children		−0.0930	−0.0846	−0.138		−0.0478	−0.0497	−0.0338
		(0.115)	(0.116)	(0.116)		(0.188)	(0.187)	(0.187)
Spouse		0.153	0.121	0.125		0.0663	0.0645	0.115
		(0.120)	(0.122)	(0.122)		(0.185)	(0.188)	(0.190)
Household size		−0.000693	−0.00883	0.00265		0.0940	0.0954	0.134 *
		(0.0487)	(0.0493)	(0.0506)		(0.0672)	(0.0724)	(0.0755)
Working as a public official			−0.151	−0.135			−0.00533	−0.0503
			(0.176)	(0.180)			(0.467)	(0.447)
Log of household income			0.0396	0.00307			−0.0142	−0.0825
			(0.0644)	(0.0639)			(0.119)	(0.121)
Log of household assets			0.0787 **	0.0446			0.0168	−0.0149
			(0.0331)	(0.0357)			(0.0465)	(0.0492)
Subjective health				0.237 ***				0.0883
				(0.0455)				(0.0542)
Future anxiety				−0.0833 **				−0.160 ***
				(0.0404)				(0.0588)
Myopic view of the future				0.00497				0.0705
				(0.0452)				(0.0598)
Constant	−9.187	−9.567	−8.816	−9.458	6.839	6.655	6.989	4.590
	(8.443)	(8.473)	(8.517)	(8.807)	(13.33)	(13.33)	(13.50)	(13.74)
Observations	1263	1263	1263	1263	631	631	631	631
Log-likelihood	−576.7	−575.7	−571.1	−549.9	−308	−306.6	−306.5	−299.6
Chi2 statistics	8.158	9.707	19.53	55.67	3.688	6.672	6.890	20.18
p-value	0.1478	0.2862	0.0523	0.0000	0.5952	0.5724	0.8079	0.1247

Note: *** p < 0.01, ** p < 0.05, * p < 0.10.

provide subsample estimation by age group. Table 7 shows the results for the participants aged both 35 years and below and between 35 and 50 years. Table 8 shows the results for the participants aged between 50 and 65 years.

Table 7 and Table 8 show that the results are largely different for the youngest cohorts and others. Although having a spouse had a significant effect on the youngest cohort, it had an insignificant effect on satisfaction among the older cohorts. Only public officials from the youngest cohort were more likely to be satisfied with the special cash payment policy. The effect of household assets was positive and consistent only among the youngest cohort. In contrast, respondents in most cohorts who stated that they were healthy were more likely to be satisfied with the policy. Subjective health had only an insignificant effect on policy satisfaction among the oldest cohort.

4. Discussion

Our results show that people were generally dissatisfied with the one-time cash payment policy (Table 2) and that there is heterogeneity in policy satisfaction among respondents (Table 5); therefore, our assumptions regarding the effectiveness of the cash payment policy may be true. Specifically, although some people may feel satisfied with the one-time cash payment policy, others may find it insufficient to relieve them of the economic hardships they have faced as a result of COVID-19. Thus, the government should consider tailoring future cash payment policies based on the needs of different groups. According to the full model specification (Model 4, Table 5), several significant variables are associated with policy satisfaction. Some are consistent with the narrative of previous studies. Some are not.

Kikuchi et al. [50] argued that female workers are more likely to experience economic hardship during the pandemic. However, our results (Model 4, Table 5) show that males were less likely to feel satisfied with the one-time payment policy. This dissatisfaction may stem from both traditional gender roles that persist in Japanese households, where male partners are expected to be breadwinners [66,67,68,69], and the loss of supporting income from female partners during the pandemic. The redundancy of non-regular jobs was expected during the pandemic and those who mainly held these jobs were women; therefore, they were more likely to lose their jobs during the COVID-19 pandemic [50,70]. This coincides with our findings on our female observations. We found that women who had more stable statuses, such as being married or working as public officials, were more likely to be satisfied with the policy (Model 4, Table 6). Men with higher skills and experience (i.e., having higher education attainment or being older) were more likely to bear more breadwinners’ mental burdens. Thus, they were less likely to feel satisfied with the one-time cash payment (Model 4, Table 6). The findings vis-à-vis the effect of future anxiety, which only existed among men (Model 4, Table 6), also support this argument.

The older population was more likely to become seriously ill from COVID-19 than their younger counterparts; therefore, the one-time payment policy was insufficient to alleviate their anxiety. Thus, older peoples were less likely to be satisfied with the one-time payment (Model 4, Table 5). The findings on the effect of future anxiety (Model 4, Table 8) also supported this argument. Future anxiety did not affect policy satisfaction among the younger peoples (Model 4, Table 7). The older peoples experienced the effect extensively (Model 4, Table 8). These findings are consistent with those of Kaneda et al. [47], who found that older peoples promptly increased their consumption after receiving the one-time payment. Therefore, they may increase their consumption in order to alleviate their anxiety.

For financial security variables, people with higher asset values, working as a public official, or having a partner were more likely to feel satisfied with the policy (Model 5, Table 7). Our subsample analysis showed that these effects only existed among the youngest peoples. The younger population was intrinsically less likely to have substantial financial burdens than their older counterparts; therefore, receiving the one-time payment while having such financially secure characteristics made them better off. This finding is consistent with those of Kubota et al. [43], who found that people with more valuable financial assets were less likely to spend their cash transfers, and with those of Kaneda et al. [47], who found that the younger peoples were less likely to spend their cash transfers. For subjective health variables, because healthy people are less likely to experience COVID-19-related hospitalization than others, people with high subjective health were less likely to have anxiety about the pandemic. Thus, a one-time payment would effortlessly compensate healthy people’s economic loss.

Our results show that the “one size fits all” policy may not be appropriate for alleviating people’s economic loss during the pandemic. Similarly to Kadoya et al. [40], our results highlighted the heterogeneity in policy satisfaction and many effects across subsamples. We found that this policy substantially helped the young population overcome their economic losses and anxieties as a result of COVID-19. On the other hand, it did not significantly help the older population. Therefore, if the government is considering implementing a similar policy, it should consider adjusting such a policy to suit the needs of each socioeconomic group. Without a tailored policy, the government may not be able to help vulnerable groups efficiently and effectively.

Our study may suffer from exclusion bias. Although we obtained 17,463 observations in 2020, the lack of responses in 2021 and missing information led us to exclude many of our observations. Our results may be biased because of the reliance on Internet surveys. According to the Ministry of Internal Affairs and Communication [71], there is an asymmetric distribution in Internet penetration rates across different household incomes [71]. Thus, our sample’s annual household income was slightly higher than an average Japanese household by around JPY 30,000 (equivalent to USD 261.84 as of 18 January 2022) [72]. Our average sample appeared older, and our sex ratio was more skewed than the national average [73]. Regardless, our study offers robust results and a consistent narrative with its Japanese counterparts. Future research should focus on a more objective measure of satisfaction with the government policy and how people of various socioeconomic backgrounds divide stimulus money in consumption and savings so that the government can effectively formulate economic revival policies.

5. Conclusions

The COVID-19 pandemic has disrupted household finance in all over the world. To alleviate the economic shock, the Japanese government distributed a one-time payment of JPY 100,000 to all its registered residents. However, hardly any studies in Japan have explored people’s satisfaction with this kind of stimulus policy implemented by the government. Thus, our study contributes not only to understanding satisfaction with a particular stimulus policy, but also to identifying vulnerable groups who still need government support. Against the backdrop of heterogenous effects of economic shocks on various socioeconomic groups, this study explored the relationship between socioeconomic status and satisfaction with a one-time cash payment policy in Japan. Our results show that the one-time cash payment policy was insufficient to alleviate people’s anxiety nationwide. We found heterogeneity in policy satisfaction among sub-socioeconomic groups and learned how their characteristics are associated with policy satisfaction. Although we understand the need for a prompt government response to a crisis, the government should consider implementing policies based on socioeconomic characteristics in the future. The Japanese government has been implementing various economic stimulus programs at the consumer and industry levels; thus, the results of our study suggest that the government should carefully select the target group so that stimulus money is channeled to the right place and the economy can recover from the recession caused by the COVID-19 pandemic.