Advertisement

The impact of declaring the state of emergency on human mobility during COVID-19 pandemic in Japan

Open AccessPublished:October 05, 2022DOI:https://doi.org/10.1016/j.cegh.2022.101149

      Abstract

      Background/objectives

      Japan has responded to the spread of COVID-19 through declaration of a state of emergency to regulate human mobility. Although the declaration was enforced by the government for prefectures, there is limited evidence as to whether the public complied with requests for voluntary stay at home. In this study, we evaluated the impact of declaring a state of emergency on human mobility during the COVID-19 pandemic in Japan.

      Methods

      We utilized daily human mobility data for 47 prefectures in Japan. Data were collected via mobile phone from February 1, 2020 to April 30, 2021. Difference-in-difference analysis was utilized to estimate the effects of the declaration of a state of emergency on prefectures in the Tokyo Metropolitan Area (Tokyo, Kanagawa, Saitama, and Chiba) in comparison to other prefectures where the state of emergency was first lifted (Osaka, Hyogo, Fukuoka, and Aichi).

      Results

      Human mobility was suppressed during the second state of emergency, from January 8 to March 21, 2021. However, the impact was weaker for the second state of emergency compared to the first.

      Conclusion

      In Japan, government requests for stay at home, such as the declaration of a state of emergency, were temporarily able to control human mobility. However, the second state of emergency was not as effective as the first. If additional need to regulate human mobility arises, self-restraint with stronger enforcement should be considered.

      Keywords

      1. Introduction

      The coronavirus infection (COVID-19) is caused by SARS-CoV-2, which was first identified in December 2019.
      • Huang C.
      • Wang Y.
      • Li X.
      • et al.
      Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.
      Most coronavirus infections cause only mild respiratory symptoms, but COVID-19 can progress to acute respiratory distress syndrome, which can be fatal.
      • Wu F.
      • Zhao S.
      • Yu B.
      • et al.
      A new coronavirus associated with human respiratory disease in China.
      Since the World Health Organization declared COVID-19 a global pandemic, deaths have continued to increase worldwide.

      COVID-19 Map–Johons Hopkins Coronavirus Resource Center, https://coronavirus.jhu.edu/map.html.

      Initially, there was no standard for treatment, the infectivity of SARS-CoV-2 and the route of transmission were unclear, and nonpharmaceutical interventions (NPIs) were an important means of deterring the spread of the disease.
      • Bo Y.
      • Guo C.
      • Lin C.
      • et al.
      Effectiveness of non-pharmaceutical interventions on COVID-19 transmission in 190 countries from 23 January to 13 April 2020.
      Most countries implemented NPIs, including human mobility restrictions, such as lockdowns. Recent reports have revealed that these NPIs are effective in deterring the spread of the virus.
      • Flaxman S.
      • Mishra S.
      • Gandy A.
      • et al.
      Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.
      • Koo J.R.
      • Cook A.R.
      • Park M.
      • et al.
      Interventions to mitigate early spread of SARS-CoV-2 in Singapore: a modelling study.
      • Siedner M.J.
      • Harling G.
      • Reynolds Z.
      • et al.
      Social distancing to slow the US COVID-19 epidemic: longitudinal pretest-posttest comparison group study.
      In addition, the effectiveness of early infection control via lockdown has also been demonstrated.
      • Bo Y.
      • Guo C.
      • Lin C.
      • et al.
      Effectiveness of non-pharmaceutical interventions on COVID-19 transmission in 190 countries from 23 January to 13 April 2020.
      ,
      • Oh J.
      • Lee H.Y.
      • Khuong Q.L.
      • et al.
      Mobility restrictions were associated with reductions in COVID-19 incidence early in the pandemic: evidence from a real-time evaluation in 34 countries.
      • Zhou Y.
      • Xu R.
      • Hu D.
      • Yue Y.
      • Li Q.
      • Xia J.
      Effects of human mobility restrictions on the spread of COVID-19 in Shenzhen, China: a modelling study using mobile phone data.
      • Zhang C.
      • Chen C.
      • Shen W.
      • et al.
      Impact of population movement on the spread of 2019-nCoV in China.
      In Japan, a variety of NPIs have been implemented. In particular, the slogan, “closed spaces with poor ventilation,” “crowded places with many people nearby,” and “close-contact settings such as close-range” was branded as the slogan to avoid the “Three Cs!” In addition, the government recommended that the public incorporate basic infection prevention measures into its lifestyle.

      A “new normal way of life” in anticipation of COVID-19 by MHLW https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000121431_newlifestyle.html (in Japanese).

      On the other hand, restrictions on human mobility were not “enforced” by law in alignment with the lockdowns in other countries. Rather, Japanese lockdowns were “requested” by the government through declaration of a “state of emergency” in which the government encouraged people to voluntarily stay at home. Previous studies have analyzed the effect of the declaration of states of emergency on human mobility.
      • Nagata S.
      • Nakaya T.
      • Adachi Y.
      • et al.
      Mobility change and COVID-19 in Japan: mobile data analysis of locations of infection.
      • Hiroi Y.
      A study on the effect of Japanese-style lock-dawn on self-restraint request for COVID-19 (in Japanese).
      • Chiba A.
      The effectiveness of mobility control, shortening of restaurants' opening hours, and working from home on control of COVID-19 spread in Japan.
      • Hara Y.
      • Yamaguchi H.
      Japanese travel behavior trends and change under COVID-19 state-of-emergency declaration: nationwide observation by mobile phone location data.
      • Nagata S.
      • Nakaya T.
      • Adachi Y.
      • et al.
      Mobility change and COVID-19 in Japan: mobile data analysis of locations of infection.
      In particular, researchers have analyzed the effects of these declarations in the context of different settings and circumstances, including densely populated areas, nighttime dwellings, and commuting hours. However, to our knowledge, no study has evaluated the effect of states of emergency in Japan on human mobility utilizing a quasi-experimental method.
      In Japan, it is unclear whether the reduction observed in human mobility during the states of emergency was the result of policy effects or public crisis awareness and voluntary action. When the first state of emergency was declared in April 2020, the disease picture of COVID-19 was unclear, and therapeutic agents were not yet approved.

      Clinical Management of Patients with COVID-19 -A guide for front-line health care workers Version 1 by MHLW: https://ajhc.or.jp/siryo/20200317-covid19.pdf (in Japanese).

      However, by the time the second state of emergency was declared in January 2021, the disease picture of COVID-19 had become clearer, therapeutic agents had been approved, and the standard of care had also improved.

      Clinical Management of Patients with COVID-19 -A guide for front-line health care workers Version 4 by MHLW https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000121431_newlifestyle (in Japanese).

      In addition, economic policies such as the “Go To Campaign” were implemented in September 2020 to increase human mobility by the end of the year.
      This study examines effectiveness of state of emergency declarations by way of difference-in-difference analysis (DID) in order to understand the effect of the state of emergency on human mobility, an objective of the declaration. Since this is the first time in history that the government has enforced the declaration of a state of emergency, it is important to clarify the trends of human mobility in Japan during this time period to inform the development future infection control measures.

      2. Methods

      In this study, we conducted a DID analysis of the impact on human mobility before and after the declaration of a state of emergency in prefectures in Japan utilizing LocationMind xPop by LocationMind, Inc. DID is a statistical method for quasi-experimental studies that utilizes observational data in order to estimate causal relationships.
      • Abadie A.
      Semiparametric difference-in-differences estimators.
      ,
      • Card D.
      • Krueger A.B.
      Minimum wages and employment: a case study of the fast-food industry in New Jersey and Pennsylvania.
      Recent research has utilized DID to show that nonpharmaceutical interventions can reduce interactions, infections, and deaths.
      • Fang H.
      • Wang L.
      • Yang Y.
      Human mobility restrictions and the spread of the Novel Coronavirus (2019-nCoV) in China.
      • Hsiang S.
      • Allen D.
      • Annan-Phan S.
      • et al.
      The effect of large-scale anti-contagion policies on the COVID-19 pandemic.
      • Lyu W.
      • Wehby G.L.
      Comparison of estimated rates of coronavirus Disease 2019 (COVID-19) in border counties in Iowa without a stay-at-home order and border counties in Illinois with a stay-at-home order.
      First, human mobility data were gathered for each prefecture to analyze trends during the COVID-19 pandemic. Next, we used DID to estimate the effects of NPIs among prefectures in the Tokyo Metropolitan Area (Tokyo, Kanagawa, Saitama, and Chiba) during the state of emergency in comparison to other prefectures where the state of emergency was first lifted (Osaka, Hyogo, Fukuoka, and Aichi). The two groups were similar in terms of population, population density, and geographic characteristics. DID was utilized to examine the intervention effects of the state of emergency declarations across the two groups.
      Mobility data were collected from the prefectures from February 1, 2020 to April 30, 2021. Information was collected from individual location data gathered through mobile phone applications,
      • Huang C.
      • Wang Y.
      • Li X.
      • et al.
      Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.
      provided by NTT DOCOMO, INC, with the consent of the users. The data were processed collectively and statistically in order to conceal private information. The original location data were gathered utilizing Global Positioning System (latitude, longitude) information that was sent at a frequency of every 5 min at the shortest interval and did not include identifying information that specified individuals.
      The mobility data used for the study included the daily interprefectural origin-destination mobility flow volume. Human mobility was logarithmically transformed. The value was calibrated using the national census to represent the total population of Japan. We estimated the following DID regression using prefecture level human mobility data: Let t be the date, x be the period where x=0 represents the pretreatment period and x=1 represents the posttreatment period. Y(t,x) is indicative of the human mobility at date t and during the x period. Let D(t,x) be the indicator of the state of emergency at date t and the period x, where D(t,x)=0 indicates that the state of emergency was not declared and D(t,x)=1 indicates that the state of emergency was declared. Then, the causal effect of the state of emergency α can be established following the DID forms:
      α=E[Y(t,1)Y(t,0)|D(t,1)=1]E[Y(t,1)Y(t,0)|D(t,1)=0]


      and be estimable by least squared regression, where E[A|B] is the conditional expectation of A given B. We compared the human mobility of similar prefectures, such as that of the Osaka prefecture for [Y(t,1)Y(t,0)|D(t,1)=0]and of the Kanagawa prefecture for E[Y(t,1)Y(t,0)|D(t,1)=1] because these locations could have had reasonably had similar human mobility statistics if the state of emergency had not been declared. The sample size for the study included 84 days, covering 21 days before and after the declaration of the state of emergency in each prefecture.
      The measurement targets were prefectures that had independent periods of emergency declarations and prefectures whose populations and human mobility statistics were similar. The period of emergency declaration and the target prefectures are shown in Fig. 1.
      Fig. 1
      Fig. 1Prefectures of the 2nd state of emergency in Japan.
      Next, DID analysis was conducted in the areas where the state of emergency was first lifted (Gifu, Aichi, Kyoto, Osaka, Hyogo, and Fukuoka) and where it was continued (Tokyo, Kanagawa, Chiba, and Saitama) during the second declaration of state of emergency. These data points were classified as a “stay” if the person had stayed for more than 1 h without exceeding the threshold (1 km). Therefore, there is a possibility of missing data in the case of movement outside the prefecture. Since we were only examining the effects of the declaration of the state of emergency from the perspective of human mobility, we believe that excluding movements outside of the prefecture would impact our results. Moreover, Tokyo was not included in this study because it has a particularly large population and population density, and there were no appropriate prefectures to use as comparison data. In addition, 2 × 2 table described the outcome (Average Treatment Effect) by time and treatment was shown in the supplement tables. All analyses were conducted using R version 4.1.1.

      3. Results

      Results revealed that the emergency declaration was effective in reducing human mobility with a significant difference between the prefectures where the state of emergency was continued versus the prefectures where the state of emergency ended.
      The main intervention effect evaluated in this study was the reduction of human flow due to the second declaration of emergency. The second emergency declaration was issued from January 8 to March 21 for the four Tokyo Metropolitan Area prefectures (Tokyo, Kanagawa, Saitama, and Chiba) and from January 13 to February 28 for the other six prefectures (Osaka, Hyogo, Kyoto, Aichi, Gifu, and Fukuoka) (Fig. 1). If the second declaration of state of emergency had retained its intervention effect, human mobility in the six prefectures where the state of emergency was first lifted would increase more in comparison to the Tokyo Metropolitan Area. However, DID revealed that in Osaka, Hyogo, Aichi, and Fukuoka, which are prefectures with similar population size and density as the Tokyo Metropolitan Area (Table 1),
      • Huang C.
      • Wang Y.
      • Li X.
      • et al.
      Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.
      human mobility was suppressed during the state of emergency. All results are shown in Table 2.
      Table 1Tokyo has ahigh population and population density and there are no comparable prefectures in the volume of huma mobility.
      PrefectureAreaTotal populationSizeAverage agePopulation composition ratio [by age]Population density (person/km
      • Wu F.
      • Zhao S.
      • Yu B.
      • et al.
      A new coronavirus associated with human respiratory disease in China.
      )
      Human mobility ratio before and after the 1st state of emergency (%)
      <15 years old (%)>65 years old (%)
      TokyoTokyo metropolitan14,047,59445.311.222.76402.680.7
      KanagawaTokyo metropolitan9,237,337A46.511.825.63823.288.0
      OsakaKansai8,837,685A47.111.727.64638.486.9
      AichiTokai7,542,415B,C45.613.025.31458.084.8
      SaitamaTokyo metropolitan7,344,765B46.811.927.01934.091.0
      ChibaTokyo metropolitan6,284,480C47.111.727.61218.589.9
      HyogoKansai5,465,002C,B48.012.229.3650.688.3
      HokkaidoHokkaido5,224,61449.810.732.16786.6
      FukuokaKyushu5,135,214C,B46.713.027.91029.887.0
      ShizuokaTokai3,633,20248.412.130.1467.288.5
      A = Kanagawa and Oska; B = Saitama and Aichi, Hyogo, Fukuoka; C = Chiba and Aichi, Hyogo, Fukuoka.
      ※ A DID analysis used Kanagawa, Saitama, Chiba and other prefectures with similar population and population density as controls.
      ※ The first-round state of emergency began on April 7 in Tokyo, Kanagawa, Saitama, Chiba, Osaka, Hyogo and Fukuoka, and on April 16th in other prefectures.
      ※ Hokkaido has a low population density as a prefecture, with about 40% of the population concentrated in Sapporo city.
      Table 2Results of difference-in-difference analysis for the second state of emergency.
      PrefectureEstimate95% CIp value
      Kanagawa - Osaka97.7%96.5%to98.9%<0.01
      Kanagawa - Aichi97.2%95.8%to98.6%<0.01
      Kanagawa - Hyogo96.3%95.0%to97.5%<0.01
      Kanagawa - Fukuoka96.4%95.1%to97.7%<0.01
      Chiba - Osaka97.3%96.2%to98.4%<0.01
      Chiba - Aichi96.7%95.4%to98.0%<0.01
      Chiba - Hyogo95.8%94.7%to96.9%<0.01
      Chiba - Fukuoka95.9%94.7%to97.2%<0.01
      Saitama - Osaka98.1%97.0%to99.1%<0.01
      Saitama - Aichi97.5%96.3%to98.8%<0.01
      Saitama - Hyogo96.6%95.5%to97.7%<0.01
      Saitama - Fukuoka96.7%95.6%to97.9%<0.01
      The following is a list of the prefectures that were compared in this study:
      • (1)
        Kanagawa compared with Osaka, Aichi, Fukuoka, and Hyogo
      After the state of emergency was lifted, human mobility in Osaka, Aichi, Fukuoka, and Hyogo increased, and the DIDs estimates were statistically significant at 21 days in Osaka (estimate = 97.7%; CI = 96.5%–98.9%), in Aichi (estimate = 97.2%, I = 95.8%–98.6%), in Hyogo (estimate 96.3%; CI, 95.0%–97.5%), and in Fukuoka (estimate = 96.4%; CI = 95.1%–97.7%).
      • (2)
        Chiba compared with Osaka, Aichi, Fukuoka, and Hyogo
      After the state of emergency was lifted, human mobility in Osaka, Aichi, Fukuoka, and Hyogo increased, and the difference-in-differences estimates were statistically significant at 21 days in Osaka (estimate = 97.3%; CI = 96.2%–98.4%), in Aichi (estimate = 96.7%; CI = 95.4%–98.0%), in Hyogo (estimate = 95.8%; CI = 94.7%–96.9%), and in Fukuoka (estimate = 95.9%; CI = 94.7%–97.2%).
      • (3)
        Saitama compared with Osaka, Aichi, Fukuoka, and Hyogo
      After the state of emergency was lifted, human mobility in Osaka, Aichi, Fukuoka, and Hyogo increased, and the difference-in-differences estimates were statistically significant at 21 days in Osaka (estimate = 98.1%; CI = 97.0%–99.1%), in Aichi (estimate = 97.5%; CI = 96.3%–98.8%), in Hyogo (estimate = 96.6%, CI = 95.5%–97.7%), and in Fukuoka (estimate = 96.7%, CI = 95.6%–97.9%).
      Since the state of emergency in Osaka, Aichi, Fukuoka, and Hyogo was lifted on February 28, and the state of emergency in the Tokyo Metropolitan Area lasted until March 21, we conducted a DID for 21 days before and after March 1.
      The trend of human mobility in Japan during the assessable period is shown in Fig. 2. Human mobility tends to decrease during vacation periods (such as the Obon holiday, the new year, and Golden Week). Although the first declaration of state of emergency coincided with the Golden Week holidays, a time when human mobility already tends to decrease, the effect of reducing human mobility during the period was clear (Fig. 1). On the other hand, human mobility increased during the period of the second state of emergency. This might be the reason that the first declaration of state of emergency came into effect during the Golden Week holidays and was issued nationwide, while the second was limited to prefectures with a large number of infected people.
      Fig. 2
      Fig. 2National human mobility and timing of major policy interventions in Japan.

      4. Discussion

      This quasi-experimental study analyzed the changes in epidemic dynamics before and after a state of emergency was declared between January and March 2021. We conducted DID to examine the effect of the state of emergency on human mobility in the Tokyo, Kanagawa, Chiba, and Saitama prefectures, which are the most densely populated metropolitan areas in Japan, and the human mobility of the Osaka, Hyogo, Aichi, and Fukuoka prefectures, which are areas where the state of emergency was lifted earlier. Results showed that the state of emergency was effective in reducing human mobility, with a significant difference found in the Tokyo Metropolitan Area, where the state of emergency was continued. In this study, we only examined the intervention effect of the state of emergency requested by the government and did not consider the effect of infection control measures taken by the prefectures independently. As for the impact of the lag, an increasing trend in human mobility was observed even before the state of emergency declaration was lifted. Therefore, we considered the intervention effect of the state of emergency to be clear, even though the results may have been weakened. However, although the effect of the second state of emergency declaration was statistically significant, the suppression of human mobility was slight, suggesting that the effect was already small during the second declaration.
      Although human mobility control has been shown to be associated with lower infection rates as an effective NPI,
      • Flaxman S.
      • Mishra S.
      • Gandy A.
      • et al.
      Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.
      • Koo J.R.
      • Cook A.R.
      • Park M.
      • et al.
      Interventions to mitigate early spread of SARS-CoV-2 in Singapore: a modelling study.
      • Siedner M.J.
      • Harling G.
      • Reynolds Z.
      • et al.
      Social distancing to slow the US COVID-19 epidemic: longitudinal pretest-posttest comparison group study.
      there have also been reports of economic impacts and other health hazards, such as increased suicide rates.
      • Sakamoto H.
      • Ishikane M.
      • Ghaznavi C.
      • Ueda P.
      Assessment of suicide in Japan during the COVID-19 pandemic vs previous years.
      The policies of countries around the world in response to the COVID-19 pandemic differed,
      • Huang C.
      • Wang Y.
      • Li X.
      • et al.
      Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.
      ,
      • Fang H.
      • Wang L.
      • Yang Y.
      Human mobility restrictions and the spread of the Novel Coronavirus (2019-nCoV) in China.
      with respect to NPIs also varying.
      • Bo Y.
      • Guo C.
      • Lin C.
      • et al.
      Effectiveness of non-pharmaceutical interventions on COVID-19 transmission in 190 countries from 23 January to 13 April 2020.
      ,
      • Xylogiannopoulos K.F.
      • Karampelas P.
      • Alhajj R.
      COVID-19 pandemic spread against countries' non-pharmaceutical interventions responses: a data-mining driven comparative study.
      ,
      • Brauner J.M.
      • Mindermann S.
      • Sharma M.
      • et al.
      Inferring the effectiveness of government interventions against COVID-19.
      For example, the United Kingdom and the United States implemented lockdowns to protect their populations from rapid increase in number of infections and deaths, while some countries, such as Sweden, resorted to voluntary social distancing, fearing that mandatory regulations would reduce health risks but not economic risks. Some countries resorted to voluntary social distancing out of concern that the cost would be too high.
      • Born B.
      • Dietrich A.M.
      • Muller G.J.
      The lockdown effect: a counterfactual for Sweden.
      Others, such as Germany, opted for a time-bound approach because human mobility restrictions were criticized as being unreasonably harsh, economically, given the lethality of COVID-19. While some countries adopted a lockdown policy to significantly limit human mobility, Japan requested voluntarily stay home via state of emergency declaration because the government could not legally enforce a lockdown. The United States and various European countries were “required” to stay at home (with some exceptions), while the people of Japan were given the “recommendation” to stay home, which was similar to the voluntary measures taken in Sweden.
      • Watanabe T.
      • Yabu T.
      Japan's voluntary lockdown.
      Sweden had 796,266 infected cases and 9113 deaths as of March 31, 2021, while Japan had 472,112 infected cases and 13,602 deaths,

      COVID-19 Map–Johons Hopkins Coronavirus Resource Center, https://coronavirus.jhu.edu/map.html.

      which is clearly low in terms of the population ratio. Although there are no perfect statistics for comparison, Japan's COVID-19 measures appear to have been relatively successful.
      Although there are reports on the effect of declaring a state of emergency on reducing human mobility,
      • Nagata S.
      • Nakaya T.
      • Adachi Y.
      • et al.
      Mobility change and COVID-19 in Japan: mobile data analysis of locations of infection.
      • Hiroi Y.
      A study on the effect of Japanese-style lock-dawn on self-restraint request for COVID-19 (in Japanese).
      • Chiba A.
      The effectiveness of mobility control, shortening of restaurants' opening hours, and working from home on control of COVID-19 spread in Japan.
      • Hara Y.
      • Yamaguchi H.
      Japanese travel behavior trends and change under COVID-19 state-of-emergency declaration: nationwide observation by mobile phone location data.
      • Nagata S.
      • Nakaya T.
      • Adachi Y.
      • et al.
      Mobility change and COVID-19 in Japan: mobile data analysis of locations of infection.
      it is unclear whether these effects are due to voluntary curtailment of human mobility because of the health risk of COVID-19 or because it was requested through the state of emergency. At the time of the first state of emergency, around April 2020, the clinical course, transmission patterns, and severity factors of COVID-19 were unclear, and therapeutic agents had not yet been approved.

      Clinical Management of Patients with COVID-19 -A guide for front-line health care workers Version 1 by MHLW: https://ajhc.or.jp/siryo/20200317-covid19.pdf (in Japanese).

      The high fatality rate among the elderly was apparent, and, in Japan, where the population is the oldest in the world, the health risk of COVID-19 might have been a reason for people to voluntarily isolate. However, by January 2021, when the second state of emergency was declared, the clinical and transmission patterns of COVID-19 had been elucidated, therapeutic agents had been approved, and the standard of care had improved.

      Clinical Management of Patients with COVID-19 -A guide for front-line health care workers Version 4 by MHLW https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000121431_newlifestyle (in Japanese).

      In addition, economic policies, such as the Go To Travel campaign, were implemented, and COVID-19 was no longer considered as an unknown disease. Under these circumstances, the fact that the second declaration of state of emergency impacted human mobility is an interesting result. On the other hand, although the effect of the second emergency declaration was statistically significant, the suppression of human mobility was slight, suggesting that the effect of the declaration was limited during the second round. Japan has now had up to the sixth wave of the COVID-19 pandemic, and up to the fifth wave, even the slight effect of suppressing human flow through measures such as declaring a state of emergency, combined with other NPIs, has succeeded in preventing the spread of infection. However, after the 6th wave, i.e., when the strain has repeatedly mutated to become an omicron strain and its infectivity has increased, it may become difficult to prevent the spread of the disease through voluntary restraints based on requests that are not legally binding. Although the omicron strain has a strong transmission ability and a weak virulence compared to conventional strains and previous mutant strains, it might be the time to consider the use of legal measures that would allow stronger regulation of human mobility in case a new virus with stronger transmission ability and even higher lethality emerges in the future.
      A limitation of this study is that the collection of mobile data cannot fully measure human mobility. Therefore, it is difficult to make definitive epidemiological conclusions through this study. In addition, this study compared human mobility during the declaration of the state of emergency from January 8 to March 21 after the declaration was lifted, instead of comparing it with the end of the year and the beginning of the year, when human mobility decreases due to obtain a regression. Additionally, a third state of emergency was enacted in Tokyo on April 25, 2021, and there was a clear increase in human mobility compared to the same period in 2020 (Fig. 2). Although this is an interesting trend, the period of data analyzed was specifically from February 1, 2020 to April 30, 2021, so it was not possible to study the trends of human mobility in detail after the third of the state of emergency. In Japan, the declaration of a state of emergency can be effective, but its effectiveness may be weakened with increase in the duration and the number of times it is enforced. Per results of this study, there is a difference between the rate of decrease in human mobility at the time of the first declaration (April–May 2020) and the rate of decrease at the time of the second declaration (January–March 2021), suggesting a decline in the effect of the policy over time. Further research on the effects of declaration of a state of emergency after the third declaration is indicated, and continuous evaluation will be necessary for future infectious disease control measures.
      In Japan, government requests to stay at home, such as the declaration of a state of emergency, were effective for decreasing human mobility. However, the second declaration of state of emergency was not as effective as the first, and in order to better regulate the overall mobility of people, consideration self-restraint with stronger enforcement may be indicated.

      Author contributions

      Nakamoto, Kawakami, Nojiri and Kuroki have full access to all of the data in this study and takes responsibility for the integrity of the data and accuracy of the data analysis. Concept and design: All authors Acquisition, analysis, or interpretation of data: All authors Drafting of the manuscript: All authors Critical revision of the manuscript for important intellectual content: All authors Statistical analysis: Nakamoto, Kawakami.

      Obtained funding

      None.

      Conflict of interest disclosures

      Nakamoto is an employee of Gilead, Japan. Other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Huang C.
        • Wang Y.
        • Li X.
        • et al.
        Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.
        Lancet. 2020; 395: 497-506
        • Wu F.
        • Zhao S.
        • Yu B.
        • et al.
        A new coronavirus associated with human respiratory disease in China.
        Nature. 2020; 579: 265-269
      1. COVID-19 Map–Johons Hopkins Coronavirus Resource Center, https://coronavirus.jhu.edu/map.html.

        • Bo Y.
        • Guo C.
        • Lin C.
        • et al.
        Effectiveness of non-pharmaceutical interventions on COVID-19 transmission in 190 countries from 23 January to 13 April 2020.
        Int J Infect Dis. 2021; 102: 247-253
        • Flaxman S.
        • Mishra S.
        • Gandy A.
        • et al.
        Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.
        Nature. 2020; 584: 257-261
        • Koo J.R.
        • Cook A.R.
        • Park M.
        • et al.
        Interventions to mitigate early spread of SARS-CoV-2 in Singapore: a modelling study.
        Lancet Infect Dis. 2020; 20: 678-688
        • Siedner M.J.
        • Harling G.
        • Reynolds Z.
        • et al.
        Social distancing to slow the US COVID-19 epidemic: longitudinal pretest-posttest comparison group study.
        PLoS Med. 2020; 17e1003244
        • Oh J.
        • Lee H.Y.
        • Khuong Q.L.
        • et al.
        Mobility restrictions were associated with reductions in COVID-19 incidence early in the pandemic: evidence from a real-time evaluation in 34 countries.
        Sci Rep. 2021; 1113717
        • Zhou Y.
        • Xu R.
        • Hu D.
        • Yue Y.
        • Li Q.
        • Xia J.
        Effects of human mobility restrictions on the spread of COVID-19 in Shenzhen, China: a modelling study using mobile phone data.
        Lancet Digit Health. 2020; 2: e417-e424
        • Zhang C.
        • Chen C.
        • Shen W.
        • et al.
        Impact of population movement on the spread of 2019-nCoV in China.
        Emerg Microb Infect. 2020; 9: 988-990
      2. A “new normal way of life” in anticipation of COVID-19 by MHLW https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000121431_newlifestyle.html (in Japanese).

        • Nagata S.
        • Nakaya T.
        • Adachi Y.
        • et al.
        Mobility change and COVID-19 in Japan: mobile data analysis of locations of infection.
        J Epidemiol. 2021; 31: 387-391
        • Hiroi Y.
        A study on the effect of Japanese-style lock-dawn on self-restraint request for COVID-19 (in Japanese).
        J City Plan Inst Jpn. 2022; 55
        • Chiba A.
        The effectiveness of mobility control, shortening of restaurants' opening hours, and working from home on control of COVID-19 spread in Japan.
        Health Place. 2021; 70102622
        • Hara Y.
        • Yamaguchi H.
        Japanese travel behavior trends and change under COVID-19 state-of-emergency declaration: nationwide observation by mobile phone location data.
        Transp Res Interdiscip Perspect. 2021; 9100288
        • Nagata S.
        • Nakaya T.
        • Adachi Y.
        • et al.
        Mobility change and COVID-19 in Japan: mobile data analysis of locations of infection.
        J Epidemiol. 2021; 31: 387-391
      3. Clinical Management of Patients with COVID-19 -A guide for front-line health care workers Version 1 by MHLW: https://ajhc.or.jp/siryo/20200317-covid19.pdf (in Japanese).

      4. Clinical Management of Patients with COVID-19 -A guide for front-line health care workers Version 4 by MHLW https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000121431_newlifestyle (in Japanese).

      5. https://www8.cao.go.jp/chosei/koubun/rekiren/rekishitekikinkyujitai.pdf.

        • Abadie A.
        Semiparametric difference-in-differences estimators.
        Rev Econ Stud. 2005; 72: 1-19
        • Card D.
        • Krueger A.B.
        Minimum wages and employment: a case study of the fast-food industry in New Jersey and Pennsylvania.
        Am Econ Rev. 1994; 84: 772-793
        • Fang H.
        • Wang L.
        • Yang Y.
        Human mobility restrictions and the spread of the Novel Coronavirus (2019-nCoV) in China.
        J Publ Econ. 2020; 191104272
        • Hsiang S.
        • Allen D.
        • Annan-Phan S.
        • et al.
        The effect of large-scale anti-contagion policies on the COVID-19 pandemic.
        Nature. 2020; 584: 262-267
        • Lyu W.
        • Wehby G.L.
        Comparison of estimated rates of coronavirus Disease 2019 (COVID-19) in border counties in Iowa without a stay-at-home order and border counties in Illinois with a stay-at-home order.
        JAMA Netw Open. 2020; 3e2011102
        • Sakamoto H.
        • Ishikane M.
        • Ghaznavi C.
        • Ueda P.
        Assessment of suicide in Japan during the COVID-19 pandemic vs previous years.
        JAMA Netw Open. 2021; 4e2037378
        • Xylogiannopoulos K.F.
        • Karampelas P.
        • Alhajj R.
        COVID-19 pandemic spread against countries' non-pharmaceutical interventions responses: a data-mining driven comparative study.
        BMC Publ Health. 2021; 21: 1607
        • Brauner J.M.
        • Mindermann S.
        • Sharma M.
        • et al.
        Inferring the effectiveness of government interventions against COVID-19.
        Science. 2021; : 371
        • Born B.
        • Dietrich A.M.
        • Muller G.J.
        The lockdown effect: a counterfactual for Sweden.
        PLoS One. 2021; 16e0249732
        • Watanabe T.
        • Yabu T.
        Japan's voluntary lockdown.
        PLoS One. 2021; 16e0252468