Changes in social environment due to the state of emergency and Go To campaign during the COVID-19 pandemic in Japan: An ecological study.

BACKGROUND: During the coronavirus disease 2019 (COVID-19) pandemic in Japan, the state of emergency, as a public health measure to control the spread of COVID-19, and the Go To campaign, which included the Go To Travel and Go To Eat campaigns and was purposed to stimulate economic activities, were implemented. This study investigated the impact of these government policies on COVID-19 spread.
METHODS: This ecological study included all 47 prefectures in Japan as samples between February 3 and December 27, 2020. We used COVID-19 cases and mobility as variables. Additionally, places where social contacts could accrue, defined as restaurants, companies, transportation, and tourist spots; mean temperature and humidity; the number of inhabitants in their twenties to fifties; and the number of COVID-19 cases in the previous period, which were factors or covariates in the graphical modeling analysis, were divided into five periods according to the timing of the implementation of the state of emergency and Go To campaign.
RESULTS: Graphical changes occurred throughout all five periods of COVID-19. During the state of emergency (period 2), a correlation between COVID-19 cases and those before the state of emergency (period 1) was observed, although this correlation was not significant in the period after the state of emergency was lifted (period 3). During the implementation of Go To Travel and the Go To Eat campaigns (period 5), the number of places where social contacts could accrue was correlated with COVID-19 cases, with complex associations and mobility.
CONCLUSIONS: This study confirms that the state of emergency affected the control of COVID-19 spread and that the Go To campaign led to increased COVID-19 cases due to increased mobility by changing behavior in the social environment where social contacts potentially accrue.

Introduction

Coronavirus disease 2019 (COVID-19) originated in Wuhan, China [1]. From there, it spread rapidly worldwide and was declared a pandemic by the World Health Organization in March 2020 [2]. At the beginning of February 2021, the number of infected people worldwide was >100 million [3]. In Japan, the first case was reported mid-January 2020, and the subsequent exponential increase in the number of cases led to 235,752 confirmed cases as of January 1, 2021 [4].

Because the main route of transmission of COVID-19 among humans is close contact and respiratory droplets [5,6] in not only symptomatic period but also asymptomatic and pre-symptomatic period [7], it is considered highly infectious. Thus, it is critical to reduce close contact. Many countries implemented public strategies, such as lockdowns and mobility restrictions, to stop the spread. After confirmation of the first wave of the COVID-19 outbreak, the Japanese government declared a national state of emergency, which was not a lockdown but rather a noncompulsory restriction (S1 Table), from April 6, 2020, for seven prefectures in metropolitan areas and from April 16, 2020, for the remaining prefectures across Japan [8]. They requested people to refrain from unnecessary outings, including going to restaurants, schools, and public facilities, and traveling to avoid the “Three Cs,” namely, closed spaces, crowded spaces, and close-contact settings. After the restrictions were lifted on May 25, 2020, the number of infected cases decreased temporarily [9]. Controversially, from the end of July 2020, the Japanese government conducted the Go To campaign as an economic measure to stimulate economic activities, which contributed to increasing social contacts and human mobility in Japan. The campaign included Go To Travel, from August 2020, which encouraged all residents to travel around Japan by providing discounts for travel expenses and coupons, and Go To Eat, from October 2020, which encouraged people to eat out at restaurants and bars by providing points for discounts [10]. When the Go To campaign had just started, the second COVID-19 wave occurred. Because the third big wave occurred during the New Year holidays, the campaign was suspended at the end of 2020, and the state of emergency was declared again at the beginning of January 2021.

Social distancing measure, stay-at-home orders and lockdowns implemented as non-pharmaceutical interventions in some countries contributed to the reduction of mobility [11-13], including domestic travel and outings to neighboring places (such as workplaces, restaurants, bars, and schools), which was associated with the mitigation of the short-term spread of the disease in previous studies [14-19]. However, some studies reported that mobility and social contacts increased substantially after lifting mobility restrictions and reopening economic activities [20,21]. Regarding environmental factors related to increased social contacts and human mobility in daily life, facilities such as restaurants, company premises, public transportation, hotels, and leisure spots are closed places where the likelihood of transmission is considered potentially high [22-24]. Thus, after lifting restrictions and reopening economic activities, mobility and social contacts would increase at these specific facilities and lead to rebound of transmission depending on the level of mitigation [25]. Weather conditions are also potential factors that influence human activity [26,27].

The impact of the state of emergency in Japan on controlling the spread of COVID-19 through mobility restrictions, various environmental factors that could have affected the spread of COVID-19 throughout 2020, and the implementation of the Go To campaign in the middle of the pandemic, which might have made the context more complicated, have not been investigated. Thus, this study aimed to clarify how the execution of the state of emergency and the Go To campaign impacted the spread of COVID-19 in Japan. We used graphical modeling, taking into account mobility, which is related to the number of social contacts in daily life, and climate based on data for five time periods related to changes in government policies.

Materials and methods

Study design

This ecological study was approved by the ethics committee of Juntendo University and was conducted using publicly available data. The study period was from the early stage of the outbreak to the end of 2020, when the Go To campaign was suspended. To verify the impact of the state of emergency and Go To campaign, we divided the study period into five stages, as follows: period 1, before the state of emergency was executed (February 3 to April 19, 2020); period 2, during the state of emergency (April 20 to Jun 7, 2020); period 3, after the state of emergency was lifted (Jun 8 to August 9., 2020); period 4, after initiating the Go To Travel campaign (August 10 to October 18,2020); and period 5, after initiating the Go To Eat campaign (October 19 to December 27). Then, we compared the relationships among the variables in each period. We designated each period with adjustment for the International Standard for the representation of dates and times (ISO 8601) and approximately 14 days after the beginning and end of each measure (Fig 1). The incubation period for COVID-19 is estimated as on average 5–6 days up to 14 days [6,7,28,29]. We assumed reporting delay of polymerase chain reaction (PCR) test existed, which was reported around 3 days [30] last year because of an immature system of investigation, and people could take PCR tests after presenting symptoms in accordance with the guidance in Japan. We considered the sum of time including incubation period, testing and reporting, and we applied 14 days as lag between periods. To investigate the robustness of results of this study which set 14 days as lag time between periods, we conducted sensitivity analysis applying 0-day lag and 7-day lag. The units of this analysis are the prefectures of Japan (N = 47), and their demographics are shown in Table 1.

Fig 1

Number of confirmed COVID-19 cases in Japan and government policies from February 3 to December 27, 2020.

The timing of the implementation of the state of emergency, Go To Travel campaign, Go To Eat campaign, and the five periods are shown.

Table 1

Baseline characteristics of the 47 prefectures.

Prefecture	Population	Population density (per km2)	Total COVID-19 cases (per 100,000)	Inhabitants in their twenties to fifties (per 100,000)
Hokkaido	5,267,762	66.5	245.7	468
Aomori	1,275,783	127.6	34.3	447
Iwate	1,235,517	79.4	30.7	443
Miyagi	2,292,385	314.8	91.2	490
Akita	985,416	81.8	12.7	418
Yamagata	1,082,296	114.2	34.1	436
Fukushima	1,881,981	132.8	47.6	456
Ibaraki	2,921,436	468.1	79.0	482
Tochigi	1,965,516	301.5	62.9	484
Gunma	1,969,439	302.8	111.1	477
Saitama	7,390,054	1,933.6	179.1	515
Chiba	6,319,772	1,217.9	165.9	509
Tokyo	13,834,925	6,367.8	408.8	569
Kanagawa	9,209,442	3,813.6	211.9	529
Niigata	2,236,042	174.8	22.6	453
Toyama	1,055,999	243.6	51.4	462
Ishikawa	1,139,612	270.0	90.7	474
Fukui	780,053	182.0	44.4	463
Yamanashi	826,579	180.6	62.9	468
Nagano	2,087,307	150.0	54.1	456
Gifu	2,032,490	185.9	101.5	470
Shizuoka	3,708,556	465.3	69.2	475
Aichi	7,575,530	1,457.8	206.9	518
Mie	1,813,859	306.1	67.5	477
Shiga	1,420,948	351.6	76.5	494
Kyoto	2,545,899	556.9	175.0	489
Osaka	8,849,635	4,627.8	326.9	512
Hyogo	5,549,568	647.4	168.0	487
Nara	1,353,837	358.4	134.4	464
Wakayama	954,258	193.5	62.5	451
Tottori	561,175	157.2	16.9	444
Shimane	679,324	99.4	30.2	426
Okayama	1,903,627	264.6	66.5	467
Hiroshima	2,826,858	329.6	108.2	478
Yamaguchi	1,369,882	219.5	39.1	434
Tokushima	742,505	173.9	26.3	446
Kagawa	981,280	505.6	29.3	460
Ehime	1,369,131	233.7	30.3	448
Kochi	709,230	97.1	89.4	431
Fukuoka	5,129,841	1,024.1	161.1	487
Saga	823,810	331.4	53.8	450
Nagasaki	1,350,769	317.3	43.4	435
Kumamoto	1,769,880	234.3	97.0	446
Oita	1,151,229	177.4	53.9	442
Miyazaki	1,095,903	137.5	65.2	433
Kagoshima	1,630,146	172.8	59.3	432
Okinawa	1,481,547	639.1	352.0	494

The data on the population, population density, and inhabitants in their twenties to fifties were obtained in 2020. The data of the total COVID-19 cases were accumulated over the study period, from February 3 to December 27, 2020.

Data source

We defined COVID-19 cases as cases of infection with severe acute respiratory syndrome coronavirus 2 that were newly confirmed as positive by PCR tests. The data of COVID-19 cases across the 47 prefectures from February 3 to December 27, 2020, were obtained from the Toyo Keizai Online “Coronavirus Disease (COVID-19) Situation Report in Japan” by Kazuki Ogiwara [31]. The passengers of the Diamond Princess cruise ship that docked in Kanagawa, travelers quarantined in airports, and returnees on government charter flights from Wuhan were excluded. We also obtained public data as follows: demographic data (age) from the Ministry of Internal Affairs and Communications (2020), meteorological data (temperature and humidity) from the Japan Meteorological Agency (2020), data on shops and facilities where social contacts could accrue from the TownPage database (2020) established by the Nippon Telegraph and Telephone Corporation (NTT, http://itp.ne.jp/), and human mobility data from LocationMind xPop. LocationMind xPop uses aggregated people flow data originally collected by NTT Docomo, Inc., through their application service “Docomo Map Navi” using only the cell phone’s location data collected upon user consent to the service’s auto GPS function, and then processed by NTT Docomo in entirety and statistically before being provided to LocationMind Inc. The original location data is GPS data (latitude, longitude) sent at a frequency of every 5 minutes at the shortest interval and does not include information that specifies individuals.

Variables

We defined twelve variables (Table 2). Eight variables (COVID-19 cases, inhabitants in their twenties to fifties, mean temperature, mean humidity, mobility from urban areas, mobility from rural areas, mobility from prefectures with ordinance-designated cities, and mobility from Tokyo) were extracted from the data source for each period. The other four variables (number of restaurants, companies, transportation systems, and tourist spots) were commonly used in the analysis across all periods. These variables were obtained from the TownPage database (geographical maps of their distribution; S1 Fig) and were selected as potential environmental factors that were considered as hotspots of COVID-19 transmission by increased social contacts and mobility [32]. These variables were considered to reflect the sociographic characteristics of each prefecture in Japan. Therefore, we used meteorological variables, inhabitants in their twenties to fifties, and the variables obtained from the TownPage database as covariates. Because the spread of COVID-19 during the previous period was considered to affect the next period, we also considered this as a factor. We selected the population in their twenties to fifties as a demographic factor because these subjects were of working age and considered more active than other generations in Japan. In fact, the cumulative number of infected cases in this age group was larger than that in other age groups in Japan [33].

Table 2

Description of the variables.

Variable	Description
COVID-19 cases (per 100,000)	Newly confirmed SARS-CoV-2 cases by PCR tests
Inhabitants in their twenties to fifties (per 100,000)	Total number of inhabitants in their twenties to fifties
Restaurants (per 100,000)	Number of restaurants falling under the category of “Gourmet and Restaurants” in the NTT TownPage database, except for bento shops
Companies (per 100,000)	Number of companies falling under the category of “Business” in the NTT TownPage database
Transportations (per 100,000)	Number of railway stations, buses, ferries, and airports falling under the category of “Life” in the NTT TownPage database
Tourist spots (per 100,000)	Number of tourist information centers, rest stops, and hot springs falling under the category of “Travel and Accommodation” in the NTT TownPage database
Mean temperature (°C)	Averaged temperature
Mean humidity (%)	Averaged relative humidity
Mobility from urban areas (people)	Total volume of human mobility from prefectures with ordinance-designated cities and from Tokyo to all prefectures except for the origin
Mobility from rural areas (people)	Total volume of human mobility from prefectures not categorized as urban areas to all prefectures except for the origin
Mobility from prefectures with ordinance-designated cities a (people)	Total volume of human mobility from prefectures with ordinance-designated cities to all prefectures except for the origin and Tokyo
Mobility from Tokyo a (people)	Total volume of human mobility from Tokyo to all prefectures except for Tokyo

SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; PCR, polymerase chain reaction, NTT, Nippon Telegraph and Telephone Corporation.

aMobility from prefectures with ordinance-designated cities and mobility from Tokyo were used instead of mobility from urban areas in the influence analysis.

We measured the volume of human mobility with a 7-day moving average and added inflow from other prefectures to the number of subjects in a prefecture(weekly changes shown in S3 Fig). Depending on the prefecture of origin, we categorized human mobility as either mobility from urban areas or and mobility from rural areas.

The population density of Tokyo was the highest, and the volume of human movement out of Tokyo was considered much higher than those from other prefectures (Table 1). Additionally, Tokyo had the highest number of infected cases among all the prefectures (Table 1). Therefore, we focused on human movement from Tokyo and conducted an influence analysis to clarify the impact on the spread of COVID-19 by excluding Tokyo from the analysis and adding two variables, namely, mobility from Tokyo to other prefectures and mobility from prefectures with ordinance-designated cities, instead of mobility from urban areas. We also calculated temperature and humidity using an averaged 7-day moving average for each day and averaged them for each period. We converted all variables, except mobility, into values per 100,000 people in each prefecture and transformed all variables, except mean temperature, into natural logarithms to adjust for the normal distribution.

Statistical analyses

We performed graphical modeling to analyze the association of COVID-19 spread with the variables in each period [34,35]. Graphical Modeling is one of powerful tools to visually analyze the conditional independence/dependence structure of the whole set of observed variables (twelve variables in the paper). The regression analysis is unsuitable to conduct such analysis because the conditional independence/dependence relationships between a response variable and covariates are a main interest of regression analysis, but those between covariates are not. In this analysis, considering what variables have edges with the infectiousness number indicates that the infectious number is the outcome variable, and the other variables, including mobility, are explanatory variables in the regression analysis. We categorized the variables into two groups, taking into account the time order: (1) the group of factors (number of restaurants, companies, transportation systems, and tourist spots; number of inhabitants in their twenties to fifties; mean temperature and humidity; and number of COVID-19 cases from the previous period applied for periods 2–5) and (2) the group of outcomes (number of COVID-19 cases and all variables of mobility). Then, we conducted recursive covariance selection for each group by means of backward elimination.

To verify the performance of the model in the selection, we applied the Goodness of Fit Index (GFI) and deviance. We determined the model when the model criteria GFI is small to some level. Detailed methodology is presented in the Supporting information (S1 Text).

The graphs were quantified by partial correlations. Statistical analyses were performed using JUSE-StatWorks/V5[36].

In this study, the graphs were interpreted in a joint distribution as (1) an undirected edge (-) between two variables, representing a correlation without order or (2) no undirected edge (-) between two variables, representing conditional independence given all other variables without directed edges. Then, we compared the structure of correlations among variables for each graph throughout the five periods, focusing on the following correlations: among the group of factors, between the group of factors and the group of outcomes, and among the group of outcomes. Finally, we verified their changes and assessed whether the government policies had impacted the spread of COVID-19.

Results

Main analysis

We confirmed 219,789 infectious cases of COVID-19 across Japan in our study cohort from February 3 to December 27, 2020. S2 Fig shows geographic maps with visual representations of changes in cumulative confirmed COVID-19 cases in each period. Fig 2 shows the estimated conditional independence structure in each period. The structure of the association between variables categorized in the group of factors (number of restaurants, companies, transportation systems, tourist spots, and inhabitants in their twenties to fifties) remained unchanged throughout the five periods. S2 Table shows the process of recursive covariance selection in graphical modeling of period 1 in main analysis. In period 1 (deviance = 16.79, degrees of freedom = 26, GFI = 0.96), inhabitants in their twenties to fifties was directly correlated with COVID-19, and strongly with tourist spots (a). Tourist spots were correlated with mobility from rural areas (b). It showed that COVID-19 was indirectly correlated with tourist spots and mobility from rural areas (Fig 2A). In period 2 (deviance = 24.47, degrees of freedom = 34, GFI = 0.93), the number of COVID-19 infections in period 1 was strongly correlated with COVID-19 in period 2 (c) (partial correlation = 0.65), (Fig 2B). In period 3 (deviance = 19.87, degrees of freedom = 29, GFI = 0.94), the direct correlation between COVID-19 in periods 2 and 3 was not significant. Inhabitants in their twenties to fifties were directly correlated with COVID-19 in period 3 (d) and with mobility from urban areas (e) (partial correlation = 0.39 and 0.23) (Fig 2C). It showed that COVID-19 was indirectly correlated with mobility from urban areas. In period 4 (deviance = 19.27, degrees of freedom = 33, GFI = 0.97), COVID-19 in period 4 was directly correlated with COVID-19 in period 3 (f) (partial correlation = 0.41). COVID-19 in period 3 was directly correlated with companies (g). It showed that companies were indirectly correlated with COVID-19 in period 3 (Fig 2D). In period 5 (deviance = 13.72, degrees of freedom = 26, GFI = 0.96), the direct correlation between COVID-19 in period 4 and COVID-19 in period 5 was not significant. COVID-19 in period 5 was strongly correlated with mobility from rural areas (h) (partial correlation = −0.43). The graph for period 5 was the most complicated. All variables related to increased numbers of social contacts (restaurants, transportations, companies, and tourist spots) were directly correlated with COVID-19 in period 5. The direct correlation of mobility from urban areas with transportations was stronger than previous period. Furthermore, the correlation of mobility from rural areas with transportations and with tourist spots were also stronger than the previous period (Fig 2E).

Fig 2

Graph in each period of the main analysis using graphical modeling.

Fig 2A–2E show graphs of period 1, 2, 3, 4, and 5, respectively. The variables in the group of factors are shown as circles on a gray background, and variables in the group of outcomes are shown as circles on a white background. The directed edge (-) from the variables in the group of factors to the variables in the group of outcomes indicates the time order.

Influence analysis

We aimed to clarify the impact of the mobility from Tokyo on the spread of COVID-19 by performing an influence analysis. The graphs were slightly complicated overall in comparison with the results of the main analysis (Fig 3 and S2 Text). Similar to the results of the main analysis, the correlation between COVID-19 in the present period and that in the previous period was only observed in periods 2 (a) and 4 (b), and COVID-19 in period 5 was strongly correlated with mobility from rural areas (c) (partial correlation = −0.41). Inhabitants in their twenties to fifties were correlated with mobility from Tokyo in all five periods (partial correlation = 0.45, 0.55, 0.48, 0.52, and 0.40, respectively) and with COVID-19 in the present period in periods 1,3–5 (partial correlation = 0.28, 0.23, 0.31, and 0.29, respectively) (Fig 3). Thus, mobility from Tokyo was not directly correlated with COVID-19 in the present period but indirectly correlated through inhabitants in their twenties to fifties. This relationship was not observed among other mobility variables, inhabitants in their twenties and fifties, and COVID-19 in the present period in period 2–5.

Fig 3

Influence analysis graphs for each period using graphical modeling.

Fig 3A–3E show graphs of period 1, 2, 3, 4, and 5, respectively. The variables in the group of factors are shown as circles on a gray background, and variables in the group of outcomes are shown as circles on a white background. The undirected edge (-) from the variables in the group of factors to the variables in the group of outcomes indicates the time order.

Primary two changes in the graphs

Two primary changes were represented in the graphs throughout the five periods. First, in the period of the state of emergency (period 2; Fig 2B), a correlation of COVID-19 cases in the present period with that in the previous period was observed, although this correlation was not significant in the period after the state of emergency was lifted (period 3; Fig 2C). Second, in the period of the Go To Travel and Go To Eat campaigns (period 5; Fig 2E), this correlation was also not significant. The potential places related to increased social contact was correlated with COVID-19 in period 5, and mobility from rural areas was correlated with COVID-19 in period 5. Additionally, similar results were confirmed in the influence analysis (Fig 3). COVID-19 cases was indirectly correlated with mobility from urban areas in some periods of the main analysis and with mobility from Tokyo in all periods of the influence analysis through inhabitants in their twenties to fifties.

Sensitivity analysis

We performed sensitivity analysis with 0-day lag and 7-day lag (S2 Text, S3 Table, S4 and S5 Figs). The structure of each graph was not completely the same as that of main analysis through all periods, but the primary two changes in the graphs were consistent with those of main analysis.

Discussion

Our study shows changes in the structure of association among the variables in the group of factors and those in the group of outcomes and changes in the entire structure of graphs throughout the five periods. The fixed structure of the association of variables (potential places related to the increase in the number of social contacts and inhabitants in their twenties to fifties) might show the structure where working generations congregate or interact in daily life. Based on this structure, the primary two graphical features and changes described in our results can be interpreted as follows.

First, the result that the strong association between COVID-19 cases in the present period and those in the previous period was not significant after the state of emergency was lifted (period 3) might be because people who contracted COVID-19 in the previous period did not affect the spread that was attributed to the effect of the state of emergency on controlling the spread of COVID-19, with a lag time. Second, the graph for period 5 was the most complicated because there were correlations among the variables related to increases in the number of social contacts, mobility, and COVID-19. Additionally, the strong association between mobility and COVID-19 in the present period was only observed in period 5. This suggests that the implementation of the Go To Travel and Go To Eat campaigns contributed to COVID-19 spread. Because the Go To campaign might have stimulated human mobility not only neighbor outings but also trips to other prefectures, the spread of infections might depend on increased mobility in the specific places such as transportations, tourist spots and restaurants. Accordingly, COVID-19 cases in the previous period (period 4) had little impact. Although the Go to Travel campaign started before period 5, campaign respondents increased considerably in October and November 2020 [37], which was within period 5.

A Japanese study also reported that travel-related increases of infectious COVID-19 cases were confirmed during the initial period of the Go To Travel campaign [38]. Although our study investigated the entire period of the execution of the campaign, the suggestion was consistent. Similarly, a previous study reported that the Eat Out to Help Out [39] scheme, which was implemented in the UK during the summer of 2020 as an economic measure to revitalize the food industry, might have caused an increase in new COVID-19 cases [40].

Furthermore, the result that mobility from Tokyo was indirectly correlated with COVID-19 through inhabitants in their twenties to fifties (Fig 3) imply that mobility from Tokyo to other prefectures especially in these generations could have had a greater indirect impact on COVID-19 cases than the mobility from other areas except for the period of the Go To Travel and Go To Eat campaigns (period 5), because this relationship was not observed in mobility from other areas.

Some studies that included Tokyo, Japan, have also suggested that behavioral changes related to eating out and traveling might have varied at the level of the individual [17] during the pandemic due to individual factors, such as social anxiety and dread, with self-restriction underlying an individual’s level of risk perception [17,41-45]. In our study, variations among the models might also imply that changes in behavior were observed throughout the periods affected by the implementation of government policies.

Findings

Changes in the conditional independence structure of variables in the five periods indicate that even though the state of emergency was a noncompulsory measure, it could contribute to the reduction of COVID-19 cases, and the Go To Travel and Go To Eat campaigns, which would increase mobility across Japan, might have led to increases in COVID-19 cases, thus influencing social environment and human behavior. Additionally, mobility from Tokyo could have had a greater impact on the spread of COVID-19 than mobility from other areas of Japan.

Limitations

This study had some limitations. First, this was an ecological study, and ecological results shown at the aggregation level might differ from those obtained at the individual level, caused by ecological fallacy. Second, we did not take the intercorrelation among the samples (prefectures) into account when constructing the graphical model. Third, we did not calculate confirmed COVID-19 cases adjusted by the number of PCR tests per population of each prefecture and the difference in the implementation of PCR tests among regions. Fourth, the possible biases from the datasets exist in the following points; we applied TownPage datasets extracted in Aug 2020, however, the number of facilities might be slightly changed through the study period due to closing or new opening; and the mobility data is estimated volume calculated from GPS data. Fifth, regarding the assessment of the impact of the government policies on the control of the COVID-19 pandemic, we did not compare situations where the government policies were or were not implemented; therefore, we might have overestimated the impact. However, there is, in fact, no means to perform such an assessment.

Conclusions

Our study suggests that the state of emergency reduced not only the influence of the number of COVID-19 cases in the prior period but also other variables, changing the social environment. Additionally, the implementation of economic measures that would increase social contacts and mobility during pandemics needs to be cautiously deliberated. As previous studies suggested, social distancing measures were considered still essential, especially in specific places right after lifting mobility restrictions and reopening economic activities during pandemic [25,46,47].

Although our study results may be as expected, this is the first study to demonstrate the effectiveness of considering various social and environmental factors to clarify whether public health and economic measures impacted the spread of COVID-19. Thus, our findings have identified important insights that should be deliberated to ensure optimal public health and economic measures in the future. Because the study period was limited to the first year of the COVID-19 pandemic, further research investigating the long-term data and mitigating the limitations of the present study are expected to contribute to a more efficient public health strategy.

Geographical maps of total number of variables from The TownPage dataset: Restaurants, companies, transportations, and tourist spots in the 47 prefectures (as of Aug 2020).

(TIF)

Click here for additional data file.

Geographical maps of the cumulative COVID-19 cases in the 47 prefectures in each period (February 3 to December 27, 2020).

(TIF)

Click here for additional data file.

Weekly total volume of inter-prefecture mobility from February 3 to December 27, 2020.

(TIF)

Click here for additional data file.

Graph in each period of the sensitivity analysis (0-day lag) using graphical modeling.

(TIF)

Click here for additional data file.

Graph in each period of the sensitivity analysis (7-day lag) using graphical modeling.

(TIF)

Click here for additional data file.

Initial nationwide public health measures during the COVID-19 pandemic in Japan and other countries.

(DOCX)

Click here for additional data file.

Process of recursive covariance selection in graphical modeling of period 1 in main analysis.

(PDF)

Click here for additional data file.

The result of statistics in model selection in sensitivity analysis.

(DOCX)

Click here for additional data file.

Detailed methodology of recursive covariance selection in graphical modeling.

(DOCX)

Click here for additional data file.

Results of the influence analysis.

(DOCX)

Click here for additional data file.

The result of sensitivity analysis.

(DOCX)

Click here for additional data file.

9 Nov 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Michele Tizzoni

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ.

4. We note that Figure S1 in your submission contain [map/satellite] images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

a) You may seek permission from the original copyright holder of Figure S1 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b) If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper investigates the impact on COVID-19 spreading of the interventions implemented by the Japanese Government to stimulate the internal economy during 2020. Graphical modeling is used to understand the relationship between a set of variables (including mobility, socio-demographic factors, and weather information) and the number of cases reported in different prefectures of Japan. The methodology is applied to 5 different periods of 2020 to study the variation in the relationships between the features studied and the impact on COVID-19 spreading of the government interventions. The natural experiment provided by the implementation of the policies described is undoubtedly interesting and worth studying. Also the approach proposed by the authors is interesting and seems suitable to the purpose. Nonetheless, I think the paper needs substantial changes before being considered for publication. Below, I provide a list of the points that I think should be addressed by the authors. I will cite the line of the paper as L followed by the number of the line I’m referring to.

Major issues

1) The Introduction explains very well the interventions implemented by the Japanese government to stimulate the internal economy. However, I think a proper framing within existing literature is missing, apart from a couple of very general sentences. Previous influential studies dealt with similar problems (i.e., understand the role of certain points of interest such as restaurant etc in the spreading), such as:

https://www.nature.com/articles/s41586-020-2923-3

https://www.nature.com/articles/s41562-020-0931-9

On a more general basis, there is a vast literature on the impact of Non-pharmaceutical interventions (especially those impacting mobility) on COVID-19 spread, here a review that can help navigating it:

https://pubmed.ncbi.nlm.nih.gov/33612922/

More work have to be done to place the work in the existing literature

2) L103, authors state that the incubation period of COVID-19 is 14 days. This is wrong, also the references provided in the paper contradicts this point. Indeed, the average incubation period for COVID-19 (pre-Delta strains) is somewhere between 3 and 5 days. This value is used to define the 5 periods of the analysis, therefore this should be revised according to the right estimate.

Nonetheless, I am not sure if the incubation period is the right quantity to define the ‘lag’ between periods (by ‘lag’ I mean the number of days that are considered as a buffer between periods, and that is now set to 14 days). Indeed, there may also be a delay linked to the reporting of cases. For example, a person infected today will start to present symptoms after the incubation period but may be notified after because of delays associated with the surveillance system. Therefore, besides choosing an appropriate value I think it is essential that authors present as Supplementary Information some sensitivity analysis of the results for different lag values (what happens to the inferred graph if we introduce no lag? Are they stable to this change? What if the lag is equal to the incubation period? What if we also account for delays in reporting? Using relatively small lags should not affect the results much).

3) The definition “places where contacts could accrue” is vague even though it is a pretty important point for the paper. How are these places selected? I see that in Table 2 some categories of places are mentioned such as “Business” and “Life”. Why have some categories been chosen and not others? Within a given category, how do you choose some places and not others? The choice of places considered hotspots for COVID-19 transmission should be supported by evidence and references.

4) The paper lacks a proper materials and methods section. This undermines the overall understanding. First, the datasets used should be better explained:

- The TownPage dataset is used to define places where social contacts could accrue. Is it open access? Is it accurate? Is it complete? How is the dataset built? What are the categories listed and the type of places? Can you provide a plot showing the distribution of different points of interest in the prefectures? Please discuss

- The LocationMindxPro is used to measure mobility variation. Is it public? Where can the data be found? What is the size of the population over which mobility is computed? Can you show a plot with (even aggregate) mobility trends in Japan during the period of study? Do you work with raw data? If so, do you apply any preprocessing/inclusion criteria? Please discuss

- Also, what do you mean exactly by mobility? Do you assign each user a home location and see whether it is seen in other prefectures during the periods? Or do you define trips? If so, what is the duration of the trip? I think more detail on how you deal with mobility data is needed

Second, also the methods are not properly described:

- What is “recursive covariance selection for each group by means of backward elimination”? (L193) I understand this is the method used to infer the graphs, but how does it work? I think a few more details would help the reader to understand the methodology

- What are the Goodness of Fit Index (GFI) and deviance? (L194) How are they defined and what is their interpretation?

- L188, is the definition of the group of factors and of outcomes something that influences the graphical modeling or is it just a definition from the authors? Why is Mobility an outcome and not a factor influencing COVID-19 cases?

Minor issues

1) the very first lines (L45-50) of the paper needs some basic citations (the origin in Wuhan, the declaration of Pandemic by the WHO, the number of REPORTED cases by February 2021)

2) At L52 authors state that SARS-CoV-2 is highly infectious because of its route of transmission. This is not entirely true, it is highly infectious also because of other reasons, such as asymptomatic and pre-symptomatic transmissions

3) L142, how do you distinguish urban and rural areas?

4) L175, I do not understand the need to first compute a seven-days moving average for humidity and temperature and then an average over the all period. Why not directly average over the period?

5) L177, why is mobility not expressed per 100,000 as other variables?

6) L165, why add the number of inflow people to the people in the prefecture?

7) I find the results very difficult to read, now the section is just a long list of sentences like this one “inhabitants in their twenties to fifties showed a partial correlation with COVID-19, having a partial correlation with tourist spots, which was correlated with restaurants and variables of mobility?” Besides just commenting on the graphs I think the authors should also make an effort to provide some interpretation behind this or use additional plots to help the reader follow the logic.

8) L334, “we did not take the intercorrelation among the samples (prefectures) into account when constructing the graphical model” How can this affect the results?

9) Among the limitations, also the possible biases coming from the datasets (especially mobility ones) used should be discussed

Reviewer #2: The study investigates the effects of the state of emergency and go-to travel policies on the number of COVID-19 cases in Japan. This study concludes that the state of emergency affected the control of COVID-19 spread and that the go-to campaign increased the number of cases by changing the locations of social contacts. The objective of the study is well motivated -- to provide evidence for lockdown policies based on big mobility data, which could be useful for policy makers. However, the statistical analysis (which the methods are very unclear) are very weak and lacks the rigor to reach any conclusions on the impact of the policies on the number of infections.

Here are some comments that I recommend the authors to address:

- The methodological procedure of the analysis is very hard to follow. I strongly recommend the authors to provide more detail on the statistical analysis method they used to produce the network plots. Are we looking at just correlations between variables? or are we able to capture any causal structures between them?

- page 17 line 241 what does this mean? "The directed edge (-) from the variables in the group of factors to the variables in the group of outcomes indicates the time order." How can we tell the 'time order'? and what do you mean by 'time order'?

- Figure 2 & 3 are visually very hard to read. Please update using more high resolution figures.

- Overall, the analysis is very weak (merely captures time-lagged partial correlation only?) and thus the discussions/conclusions are not convincing. At least some econometric analysis that can give statistical conclusions on the effects of covariates on the dependent variable is needed; there are already many studies that use epidemiological models (e.g., SIR, SEIR) to rigorously achieve these kinds of analysis, which may be a path that the authors could pursue.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

10 Feb 2022

Thank you for the careful review of our manuscript.

We have considered the helpful comments from the reviewer and the editor, and we have revised the manuscript in response to these comments. The changes made are highlighted in the manuscript, and a detailed point-by-point response is described in the document " Respons to Reviewers".

We hope that, with the revisions described herein, the article is now acceptable for publication in PLOS ONE.

Thank you for taking the time to consider the manuscript. I look forward to hearing from you soon.

Submitted filename: Response to Reviewers_Kanamori et al.docx

Click here for additional data file.

8 Apr 2022

Changes in social environment due to the state of emergency and Go To campaign during the COVID-19 pandemic in Japan: an ecological study

PONE-D-21-31460R1

Dear Dr. Nojiri,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Michele Tizzoni

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: I believe all my comments have been addressed. The formatting of the manuscript is different from the standard PLoS ONE articles (Introduction, Results, Discussion, Materials and Methods) so it may be worth changing the structure of the paper, but I leave the decision to the editor.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

19 Apr 2022

PONE-D-21-31460R1

Changes in social environment due to the state of emergency and Go To campaign during the COVID-19 pandemic in Japan: an ecological study

Dear Dr. Nojiri:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Michele Tizzoni

Academic Editor

PLOS ONE