The effectiveness of vaccination, testing, and lockdown strategies against COVID-19

Fritz, Marlon; Gries, Thomas; Redlin, Margarete Article — Published Version The effectiveness of vaccination, testing, and lockdown strategies against COVID-19 International Journal of Health Economics and Management Provided in Cooperation with: Springer Nature Suggested Citation: Fritz, Marlon; Gries, Thomas; Redlin, Margarete (2023) : The effectiveness of vaccination, testing, and lockdown strategies against COVID-19, International Journal of Health Economics and Management, ISSN 2199-9031, Springer US, New York, NY, Vol. 23, Iss. 4, pp. 585-607, https://doi.org/10.1007/s10754-023-09352-1 This Version is available at: https://hdl.handle.net/10419/307055 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) International Journal of Health Economics and Management (2023) 23:585–607 https://doi.org/10.1007/s10754-023-09352-1 1 3 RESEARCH ARTICLE The effectiveness ofvaccination, testing, andlockdown strategies againstCOVID‑19 MarlonFritz1 · ThomasGries1 · MargareteRedlin1 Received: 25 November 2021 / Accepted: 17 March 2023 / Published online: 27 April 2023 © The Author(s) 2023 Abstract The ability of various policy activities to reduce the reproduction rate of the COVID-19 disease is widely discussed. Using a stringency index that comprises a variety of lockdown levels, such as school and workplace closures, we analyze the effectiveness of government restrictions. At the same time, we investigate the capacity of a range of lockdown measures to lower the reproduction rate by considering vaccination rates and testing strategies. By including all three components in an SIR (Susceptible, Infected, Recovery) model, we show that a general and comprehensive test strategy is instrumental in reducing the spread of COVID-19. The empirical study demonstrates that testing and isolation represent a highly effective and preferable approach towards overcoming the pandemic, in particular until vaccination rates have risen to the point of herd immunity. Keywords Pandemics· COVID-19· Economics· Vaccination· Testing· Nonpharmaceutical interventions· Effectiveness JEL Classification I18· C23 Introduction The global COVID-19 outbreak has created a wide range of responses from governments. The introduction and subsequent effect of government policies are subject to much debate even in recent times where most governments started to re-open and relax most of the restrictions. These policies vary between countries independently of their level of economic and social development. While the ability of lockdowns and vaccination schemes in order to curb the spread of COVID-19 is undisputed, little is known about the detailed effect of different lockdown strategies on the spread of COVID-19 and its interaction with * Margarete Redlin marg[email protected] Marlon Fritz [email protected] Thomas Gries thomas.gr[email protected] 1 Department ofEconomics, Paderborn University, Warburger Str. 100, 33098Paderborn, Germany 586 M.Fritz et al. 1 3 widespread testing. Following (Kerr etal., 2020) testing and quarantine strategies are as important as vaccinations, especially in the short run, and so the importance of a broad testing strategy is very much a subject of public debate. Furthermore, the recent development of new variants of COVID-19 and a stagnating vaccination rate also demonstrate the importance of testing and lockdown strategies in the long run, which are further discussed in the literature section. Additionally, it must be noted that population-scale testing strategies are also important because they provide important insights into the development of the pandemic (Mercer & Salit, 2021). They provide critical viral prevalence data to steer our response to the pandemic, allow measurement of the rate of virus spread in a population and helps identify regional hotspots and high-risk subpopulations. In this paper we analyze the effects of government decisions in 37 OECD countries, distinguishing between different degrees of lockdown, e.g., the shutdown of schools, restaurants, sport clubs, and other social activities. We also look in detail at the impact of vaccinations on the change in the reproduction rate. Finally, we analyze the impact of testing on the decline in the reproduction rate. Although (Ge etal., 2021) disentangle the interplay between policies and vaccination, the effect of test strategies on the spread of COVID19 remains an open topic. To shed more light on the effectiveness of testing, government restrictions, and vaccination, we simultaneously analyze the influence of all three measures on the spread of COVID-19, specifically on the reproduction rate. Thus, we extend the analysis of Ge etal. (2021) who analyze the effects of non-pharmaceutical interventions (NPI) and vaccinations by including a detailed test strategy in their investigation. Therefore, we extend the standard SIR model as described in Cherif and Hasanov (2020) or (Avery etal., 2020) by allowing for the influence of testing, lockdowns, and vaccination. We propose that a test strategy in combination with vaccination is both an effective and economically useful approach to reduce the spread of COVID-19 in the long run. The results of our baseline regression indicate that testing, vaccinations, and stricter policy measures, as expected, are associated with a significant reduction in the reproduction rate. A comprehensive test strategy is particularly effective in reducing the reproduction rate, since a one-unit increase in daily testing (e.g., from the mean value 3.22–4.22 tests per one thousand people) has a similar effect as increasing the vaccination rate by 3%. Our findings help to explain cross-country differences in COVID-19 spreading. Combined social distancing and testing strategies in the early phases of the epidemic are demonstrated to be more efficient at reducing the disease burden, and they can delay the peak of the disease. Whereas recent papers consider non-pharmaceutical interventions in isolation (Ge etal., 2021; Kerr etal., 2020), largely disregarding other instruments, our analysis looks at stringency measures in context with other policy measures and hence provides a more comprehensive picture. The detailed analysis of various components of the lockdown stringency index shows that school and workplace closures have the strongest (significant) effects. Nevertheless, they come with major social and economic costs; what is more, especially the long-run consequences of school closures and other social distancing measures are unpredictable. Therefore, for as long as vaccine doses are not available in sufficient quantity or vaccinations are not embraced by the population, a “test, trace, and isolate” strategy can help to keep the pandemic under control without having it continuing to influence our lives in unknown ways. The remainder of this paper is organized as follows. Sect.“Literature“ contains an overview of recent studies with a focus on empirical findings. Sect.“Theoretical model“ introduces the standard SIR model, which comprises policies, vaccinations, and testing strategies. Sect.“Estimation strategy“ presents the transition from theory to the empirical model. 587 The effectiveness ofvaccination, testing, andlockdown… 1 3 Sect. “Empirical evidence“ shows empirical evidence, and Sect. “Concluding remarks“ concludes. Literature There is a large and growing body of literature on the effects of COVID-19 ranging from individual health analyses to aggregate economic studies. Nevertheless, some effects remain uninvestigated. (Hartl et al., 2020) analyze the effect of lockdowns in Germany using data from Johns Hopkins University. The data are advantageous in that they match various statistics, i.e., from Germany’s Robert Koch Institute and the WHO. The authors seek to estimate the effect of different policies by identifying a trend break in the cumulated number of confirmed cases of COVID-19 after March 20, 2020. Specifically, they identify a drop in the growth rate of reported cases from 26.7 to 13.8% after March 20, 2020. The German government introduced new policies on March 13, 2020, so there is an expected 7-day lag which can be split into 5 days for the incubation period (Lauer etal., 2020; Linton etal., 2020), plus up to 2 days until testing, and an additional 1 or 2 days until the case is reported (Hartl etal., 2020). Alfano and Ercolano (2020) also argue for the effectiveness of lockdowns in reducing the reproduction rate. In contrast to Hartl et al. (2020) the authors use a panel data set with a lockdown measure for 100 countries. Their results show that lockdowns are able to reduce the number of COVID-19 cases. They distinguish between two different policy strands: (i) health policies and (ii) policies aimed at reducing the spread of COVID-19. Usually, the latter are characterized by high economic costs. Consequently, the economic debate focuses on the “trade-off between safeguarding citizens’ health and avoiding damage to the economy” [(Alfano & Ercolano, 2020), p. 510; see also Goolsbee and Syverson 2021]. However, this debate is restricted to single countries, a restriction that is relaxed within our analysis. We want to disentangle the effect of different lockdown policies to determine the optimal trade-off between safeguarding individuals and reducing the cost to the economy. Ashraf (2020) finds a similar negative effect of government-imposed social distancing on the number of confirmed cases. However, he focuses on the impact on stock markets and shows positive effects of testing and quarantining on market returns. Jarvis etal., (2020) show that lockdown policies decrease the reproduction rate from 2.6 to 0.62. Askitas etal., (2020) analyze the effects of a range of policy activities in more detail by proposing that the spread of COVID-19 is influenced by environmental, behavioral, and social factors. They investigate the effect of type and intensity of policies on daily incidence rates. Their results show that cancelling public events and imposing restrictions on private gatherings are the most efficient strategies. School closures, too, have a strong influence on the daily incidence of COVID-19. Surprisingly, workplace closures and stayat-home requirements are not significant. Cherif and Hasanov (2020) propose that besides lockdowns and isolation, testing is another economically preferable strategy for curbing the spread of COVID-19. In their theoretical study, strategic group testing and periodic testing are shown to be effective in overcoming the pandemic. Also (Baldwin, 2020) strongly recommends a massive increase in testing capacities. He points out that besides vaccinations, testing is a way “to isolate the sick from the healthy” [(Baldwin, 2020), p.1]. (Taipale etal., 2020) show that testing is effective at any stage of the pandemic and can be used in addition to other 588 M.Fritz et al. 1 3 curbing strategies such as partial lockdowns. In other words, the number of individuals who are identified and isolated matters to the development of the pandemic [(Taipale etal., 2020), p. 2]. The authors propose that ideally, almost everyone should undergo appropriate PCR testing in order to reduce the reproduction rate to below 1. Moreover, (Taipale etal., 2020) argue that a testing strategy has an advantage over a general lockdown since the population and the economy can accept many false-positive tests. The number of false-positive tests is in any case lower than the false-positive rate during a general lockdown. Although some evidence has been provided in simulation studies, empirical studies on the influence of testing on the reproduction rate are missing so far. Kerr etal., (2020) also propose to increase testing levels by conducting a simulation exercise for the Seattle metropolitan area. They find that effective isolation and routine testing are capable of reducing the reproduction rate. Nevertheless, (Kerr etal., 2020) warn that a testing strategy is only effective in curbing the spread of COVID-19 when case numbers are relatively low. The final and possibly most important driver of the reproduction rate in the long run is vaccination. Although other variants of COVID-19, as for example the Omicron variant may have an influence, they are not explicitly modelled since vaccination also weaken their effect. Although the ability of vaccinations to curb the spread of COVID-19 is undisputed, estimates concerning the threshold in order to achieve herd immunity are rather inconclusive. Ke etal., (2021) estimate the threshold for herd immunity to be between 71 and 84%. They recommend encouraging the population to get vaccinated and advocate for more public education so herd immunity can be achieved. Moreover, the study demonstrates that strong control measures are a driver when it comes to curbing the spread of COVID-19. Further (Fontanet & Cauchemez, 2020) use a more optimistic threshold by proposing herd immunity if 50% of a population are immune. Nevertheless, looking at current numbers out of the UK it is clear that this threshold is too low. Ge etal., (2021) analyze the interaction between testing and vaccination in 133 countries across different waves. Although vaccination has a growing effect in reducing the spread of COVID-19, NPIs are the most effective measures so far. Furthermore, new variants and their possible resistance to vaccinations reinforce the debate on NPIs. For example, Germany has a vaccination rate around 75% (March 2022), which makes an argument for the importance of other effective policies. Since new variants and their influence cannot be forecasted, recent literature proposes results “that policymakers and individuals should consider maintaining non-pharmaceutical interventions and transmission-reducing behaviours throughout the entire vaccination period” [(Rella etal., 2021), p.1]. Moreover, frequent large-scale testing should remain part of strategies to contain COVID-19 since it can substitute for many non-pharmaceutical interventions that come at a much larger cost to individuals, society, and the economy (Gabler etal., 2021). One constraint of existing studies is country differences, since population density, demographic factors, and weather play important roles. (Ge et al., 2021) conclude that vaccination is the most promising way out of this pandemic when accessible in all countries equally. Their study demonstrates that the cumulative effectiveness of vaccination increases. In other words, it is important to understand that individuals acquire the desired level of immunity 12days after their first dose, and that the effect of vaccination grows exponentially as herd immunity is approached. In this paper we extend the analysis of Ge etal. (2021) by including testing strategies in our investigation of NPIs and vaccination. Therefore, we extend the standard SIR model as described in Cherif and Hasanov (2020) or (Avery etal., 2020) to allow for the impact of testing, lockdowns, and vaccination. The empirical analysis rounds off our theoretical considerations. 589 The effectiveness ofvaccination, testing, andlockdown… 1 3 Theoretical model To empirically examine the effectiveness of policies that can and indeed are used to curb the spread of the disease, we first need to identify them in a theoretical model. We depart from a standard epidemiological model such as the SIR (Susceptible, Infected, Recovery) model as discussed by, e.g., (Cherif & Hasanov, 2020) or (Avery etal., 2020). In this model, individuals are in one of these three states at any given time. Given a total population of N, the number of susceptible S(t) individuals, the number of infected I(t) individuals and the number of recovered R(t) individuals adds up to S(t) + I(t) + R(t) = N. With an infection rate β and a recovery rate γ and assuming that N = 1, we can rewrite this standard SIR model of epidemic dynamics as The SIR model is a basic model. There are a number of variations and extension of the SIR model. E.g., after recovery from an infection individuals may have only temporary immune protection. In such a case the model would turn into another class of models, the SIRS model (Susceptible, Infected, Recovered, Susceptible). However, for the virus variation which we consider in this paper this scenario is seemingly not the case. Modelling policy effects ontheeffective reproduction rate An additional important characteristic of any epidemic is the basic reproduction rate RR(0), which indicates how many other people an infected individual infects under natural starting conditions. Thus, the basic reproduction rate is akin to a descriptor of the natural dynamics of a disease. In the SIR model, the basic reproduction rate is the “natural ratio” of newly infected people to one infected person per unit of time 𝛽 divided by the simultaneous recovery rate 𝛾 per unit of time Without policy interventions, a disease will naturally spread if the ratio of newly infected to recovering individuals is larger than one ( RR(0)>1 ). If RR(0)<1 , the disease will disappear because every day more individuals recover than contribute to a further spread. As we are interested in policies that curb the spread of a disease from the start, we need to develop a view of this model that allows for a discussion of effective policy measures. As the reproduction rate indicates if a disease spreads (RR > 1), stagnates (RR = 1) or declines (RR < 1), the reproduction rate is a major indicator of the epidemic dynamic. Thus, we use the reproduction rate RR to discuss the infection process and potential policy interventions in detail. To do this, we depart from the basic reproduction rate and use the effective reproduction rate. The effective reproduction rate defines a time dependent  S (t)=−𝛽S(t)I ( t ) N  I (t)=𝛽S(t) I(t) N −𝛾I(t )  R(t)=𝛾I(t). RR (0)= 𝛽 𝛾 590 M.Fritz et al. 1 3 reproduction rate when certain policies are introduced and a fraction of the host population that is susceptible is included.1 For a stylized discussion of measures to control the spread of the disease we look at the effective reproduction rate as In this equation the rate of newly infected individuals (𝛽=p𝜅∗𝜅) is explained by 𝜅, the average number of potentially infectious contacts of an infected individual, the probability of transmission with each contact p𝜅 . With pS= S(t) N we describe probability that a contact is a susceptible individual. Thus, we can rewrite With the help of this equation we can examine the effects of policy measures. In the empirical section we examine three of the following four policies: (1) distancing rules, (2) degrees of lockdowns, (3) testing and quarantine strategies, and (4) vaccination strategies. (1) To motivate the first two (and also most common) measures, we can look at the parameters p𝜅 and 𝜅 in more detail. Standard measures to reduce the probability of transmission are social distancing and protective measures such as masks. Therefore, the original probability of transmission p𝜅0 is reduced by the degree of the distance rule d(t) which holds at time t, such that we obtain Note that d(t) simple depicts the effect of distance measures on the probability of transmission. Different measure may have even different and even non-linear effects on this probability. E.g., wearing a simple textile mask may already have an effect on the transmission of the virus. However, with a medical mask this transmission probability may further decline and with an FFP2 mask it may even improve exponentially. However, the result of these various is a reduction in transmission probability which we describe with parameter d(t). (2) Another common policy measure is to reduce the number of contacts an individual can have. Without any policy, the average infected individual is characterized by 𝜅0 contacts. If the number of contacts is reduced, infected but as yet undetected individuals will have fewer contacts. Therefore, a standard policy instrument is to reduce the number of contacts via varying degrees of lockdown. Depending on the stringency of government actions at time t, the original number of contacts 𝜅0 can be reduced further by a certain percentage rate of contact reduction c(t) , such that RR (t)= p 𝜅 ∗𝜅 𝛾 pS(t ) RR (t)= p 𝜅 (t)∗𝜅(t) 𝛾 S ( t ) N p𝜅(t)=p𝜅0(1−d(t)) 𝜅(t)=𝜅0(1−c(t)). 1 See e.g. (Nishiura and Chowell 2009). 591 The effectiveness ofvaccination, testing, andlockdown… 1 3 As we will discuss in the empirical section, the reduction of contacts will be indicated by a stringency index. (3) The next measure to reduce the number of contacts that spread the disease 𝜅 is quarantining. If all individuals who are infected are immediately detected and quarantined, the number of infectious contacts reduces to zero and the disease cannot spread. However, this is not realistic, especially not in the case of COVID-19 given that some infected individuals do not show any symptoms. While symptomatic individuals can be identified and quarantined, asymptomatic individuals may not and hence continue to spread the virus. This risk can be reduced by implementing systematic testing and isolation. If an infected asymptomatic individual is tested positive and quarantined, they cannot spread the virus and hence contribute to the infection process. Therefore, the number really infected individuals that spread the disease can be reduced through this detection. How can these asymptomatic individuals who spread the virus invisibly be found? If the asymptomatic infected are distributed randomly in the total population, and 𝜏(t) is the share of the total population per period that is tested, 𝜏(t) is also the share of the asymptomatic infected individuals invisibly spreading the virus. These infected individuals can now be quarantined so they no longer contribute to the epidemic. So, the test activity reduces the total number of infectious contacts by rate 𝜏(t) —some infected individuals are detected and completely neutralized trough quarantine, and some are still fully contributing to the infectious dynamics. However, since the R value is about passing on the virus from an average infected, this effect of testing can be included in the R value as a reduction of the average number of contacts of an infected and still invisibly spreading individual. Thus, the remaining average number of contacts of an average infected and invisibly spreading individual is now Thus, a general systematic testing per period would reduce the average number of contacts of an average infected individual. (4) Vaccination is expected to not only protect the vaccinated individual, but also reduce or even completely eliminate the risk of further spreading. Vaccination affects the probability of an infected individual randomly meeting a susceptible individual, pS(t)= S(t) N . Thus, the number of susceptible individuals is reduced by the share v(t) of vaccinated individuals at t. If r(t) is the total share of recovered individuals and if we normalize the total number of individuals to one, we obtain After this short and simplifying discussion of each type of policy, we now write down the entire mechanism that determines the effective reproduction rate: 𝜅(t)=𝜅0−𝜏(t)𝜅0=𝜅0(1−c(t))(1−𝜏(t)). pS(t)=S(t)∕N=(1−v(t))(1−r(t)) RR (t)=p𝜅0𝜅0 𝛾 distance policy ⏞⏞ ⏞⏞⏞ ⏞⏞ (1−d(t)) lockdown policy ⏞⏞⏞⏞⏞ (1−c(t)) test & quar policy ⏞⏞⏞⏞⏞ (1−𝜏(t)) vaccination policy ⏞⏞⏞⏞⏞ (1−v(t))( 1−r(t) ) 592 M.Fritz et al. 1 3 If we assume that the rate of recovered individuals at the beginning of the process is close to zero, we see that the effective reproduction rate is the basic reproduction rate corrected by the policy instruments RR(t)=RR(0)(policy effect) . However, note that the four policies have different qualities. The only policy with a permanent effect after a onceand-for-all action is vaccination. The accumulated stock of vaccinations will permanently reduce the effective reproduction ratio; once herd immunity is reached, none of the other policies are needed. All other instruments are only effective for as long as they are applied. A lockdown only brings down the effective reproduction rate for as long as it remains in place. As soon as it is lifted, the infection dynamics start up again, as reflected in the basic reproduction rate. However, we also see that all policies are substitutes. A reduction in contacts per day through a lockdown is generally substitutable to systematic tests per day combined with a quarantine policy. The possibility to substitute one policy for another allows for an efficient policy choice. Thus, policy-makers can choose the cheapest policy instrument given the same reduction in infection dynamics. Estimation strategy In the previous theoretical considerations we derived how policy measures potentially affect the effective reproduction rate. With the help of this modeling we can now motivate our estimation strategy and the particular estimation models. We derived four COVID-19 policy measures that potentially affect the reproduction rate. These are: (1) distance rules, (2) degrees of lockdown, (3) testing and quarantine strategies and (4) vaccination strategies. Since the exact policy measures and expected outcomes such as the number of contacts or the distance rule are difficult to measure directly, in the empirical model we proxy these measures by available variables. First, lockdown policy is represented by the stringency index which is a lockdown composite measure based on nine government response indicators and presents the non-pharmaceutical interventions. Unfortunately, to our best knowledge, there is no data available, which would allow to measure the distance rule for our country set and daily base. However, we believe that this is also indirectly determined by the non-pharmaceutical interventions included in the analysis. Second, test and quarantine policy is proxied by the number of COVID-19 tests carried out in the population. Finally, vaccination policy is represented by the vaccination rate in the population. Our starting point is a model where we estimate a linear combination of the policy measures Δ RR(t)=𝛼+ lockdown and distance policy ⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞ 𝛽 1 stringenc index i,t−1 test & quar policy. policy ⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞ +𝛽 2 tests i,t−1 + vaccination policy ⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞ 𝛽 3 vaccinations i,t−7 +𝜇 i +𝜀 i,t where ΔRRi,t is the change in the reproduction rate in country i at day t, testsi,t−1 is the number of new COVID-19 tests per 1,000 people in country i at day t-1, vaccinationsi,t−7 , is the number of people who received a vaccination per 100 people in country i at day t − 7, stringencindexi,t−1 is a composite government response indicator in country i at day t − 1, and the disturbance term is composed of the individual effect 𝜇i and the stochastic disturbance 𝜖i,t . We select the change in the reproduction rate as the dependent variable. This allows us to measure the effect of policy measures on reproduction rate dynamics while 599 The effectiveness ofvaccination, testing, andlockdown… 1 3 therefore transform the vaccination variable and generate categorial variables based on a certain vaccinated proportion of the population, i.e. the variable Vaccinations 30 takes the value 1 if at least 30% of the population is vaccinated. The results including time fixed effects and a proportion of vaccinated of 30, 40 and 50% are shown in Table4 and confirm our previous findings. When additionally controlling for the time trend the effects of tests and state stringency stay robust in terms of effect size and significance. Furthermore, the approach with categorial vaccination variables allows us to specify the effect of vaccination in more detail. While no effect is observed for 30% of the vaccinated population, the expected negative effect is observed for a proportion of more than %. The dummy for a proportion above 50% again shows an insignificant result. However, this may be due to the very small number of cases (less than 2%) that reach this value in the observation period. Next, we examine the individual policy measures contained in the stringency index. We divide the index into its individual components and test these simultaneously. Table5 presents the results. While the test and vaccination variables remain highly significant, only three of the nine index components show significant results. School and workplace closures show the strongest effects, which are very similar in terms of size and significance. A one-unit increase in the indices—for example, from 0 (no measures) to 1 (closures recommend—is associated with a 0.4% decrease in the reproduction Table 4 Fixed-effects Results with time fixed effects and categories of vaccinated population t statistics in parentheses *p < 0.10, **p < 0.05, ***p < 0.01 (1) (2) (3) (4) (5) (6) ΔRR ΔRR ΔRR ΔRR ΔRR ΔRR Tests t-1 − 0.00041*** (− 6.126) − 0.00040*** (− 5.986) − 0.00041*** (− 6.008) − 0.00102*** (-9.789) − 0.00106*** (− 10.113) − 0.00098*** (− 9.262) Vaccinations 30 t-7 0.00133 (1.003) − 0.00029*** (− 4.811) − 0.00016*** (− 4.380) − 0.00015 (− 1.196) Vaccinations 30 (full) t-7 − 0.00034 (− 0.149) Vaccinations 40 t-7 − 0.00326** (1.965) Vaccinations 40 (full) t-7 − 0.00568** (− 1.994) Vaccinations 50 t-7 0.00174 (0.653) Vaccinations 50 (full) t-7 0.00403 (1.429) Stringency index t-1 − 0.00026*** (− 6.743) − 0.00025*** (− 6.131) − 0.00026*** (− 6.380) − 0.00037*** (− 8.839) − 0.00040*** (− 9.207) − 0.00035*** (− 8.255) Time fixe effects Yes Yes Yes Yes Yes Yes R2 within 0.068 0.088 0.069 0.089 0.068 0.089 Countries 37 36 37 36 37 37 Obs 3810 3366 3810 3366 3810 3810 600 M.Fritz et al. 1 3 growth rate. A one-unit increase in restrictions on internal movement goes along with a 0.1% decrease, respectively. This confirms previous findings that point at significant effects of non-pharmaceutical interventions like school (Banholzer et al., 2020; Ge etal., 2021; Sharma etal., 2021) and workplace closures (Askitas etal., 2021; Banholzer et al., 2020), and which argue that general lockdowns (Flaxman et al., 2020; Table 5 The Effects of COVID-19 Response Indicators t statistics in parentheses *p < 0.10, **p < 0.05, ***p < 0.01 (1) (2) (3) ΔRR ΔRR ΔRR Tests t-1 − 0.00043*** (− 6.516) − 0.00045*** (− 6.776) − 0.00044*** (− 6.570) Vaccinations t-7 − 0.00014*** (− 4.377) − 0.00015*** (− 4.874) − 0.00013*** (− 4.078) School closures t-1 − 0.00364*** (− 6.165) − 0.00367*** (− 6.174) School closures (1, 2, 3) t-1 0.00789 (1.125) School closures (2, 3) t-1 − 0.00004 (− 0.042) School closures (3) t-1 − 0.00654*** (− 7.896) Workplace closures t-1 − 0.00449*** (− 6.293) − 0.00433*** (− 6.081) Workplace closures (1, 2, 3) t-1 0.00124 (0.216) Workplace closures (2, 3) t-1 − 0.00288* (− 1.877) Workplace closures (3) t-1 − 0.00488*** (− 6.268) Cancelled public events t-1 0.00188 (1.530) 0.00112 (0.906) 0.00127 (0.992) Restrictions on gatherings t-1 0.00095 (1.160) 0.00017 (0.195) 0.00077 (0.899) Public transport closures t-1 − 0.00003 (− 0.029) − 0.00031 (− 0.276) 0.00001 (0.005) Stay-at-home requirements t-1 − 0.00032 (− 0.380) − 0.00020 (− 0.230) − 0.00005 (− 0.053) Restrictions on internal movement t-1 0.00109* (1.664) 0.00091 (1.353) 0.00090 (1.354) International travel controls t-1 0.00062 (0.662) 0.00028 (0.299) 0.00059 (0.621) Public information campaigns t-1 − 0.00330 (− 0.487) − 0.00076 (− 0.112) − 0.00217 (− 0.319) R2 within 0.042 0.049 0.043 Countries 37 37 37 Obs 3794 3794 3794 601 The effectiveness ofvaccination, testing, andlockdown… 1 3 Noland, 2021) and internal movement restrictions (Vannoni etal., 2020) reduce mobility and lower the infection rate. However, in most of these papers, non-pharmaceutical interventions were considered in isolation and other instruments were largely disregarded. By contrast, our analysis examines the stringency measures in context with other policy measures, providing a more comprehensive picture. To shed more light on the effects of the highly significant index components, we generate categorial dummies for the different values of the ordinal index scale. For example, the dummy School closures (1, 2, 3) t-1 is 1 for the categories 1 (closure recommended), 2 (closure required on some levels), and 3 (closure required on all levels), and 0 for 0 (no measures). The dummy School closures (2, 3) t-1 is 1 for the categories 2 and 3 and 0 for the categories 0 and 1. The dummy School closures (3) t-1 is 1 for the category 3 and 0 for the categories 0, 1 and 2. This allows us to identify whether the individual policy measures mapped by each level of the indices have a significant effect. The results are presented in columns two and three of Table5. While the dummies including the first and second categories are not significant, the dummy representing required closures on all levels shows a highly significant result. While the general recommendation to close or open schools with alterations and selective closures on only some levels show no significant effects, a stronger restriction in the form of required closures for all school levels is significantly related with reductions in the reproduction rate. The situation is similar for job restrictions. Again, the inclusion of the lightest restrictions in the form of a recommendation to close and to work from home shows no significant effect. Only stricter restrictions such as required closures or a firm rule to work from home for some sectors or categories of workers (category 2) and required closures for all but essential workplaces (category 3) lead to a significant correlation with a reduction in infection rates. While the dummy that includes the second category only shows significance at the 90% level, the strictest category is highly significant and shows a stronger effect in terms of magnitude. Overall, our results show that a mix of different measures is necessary to combat the pandemic. Until vaccination has progressed to the point where herd immunity has been achieved or where there is no option to achieve herd immunity, the pandemic can also be controlled by non-pharmaceutical interventions and testing, quarantine, and isolation strategies. This is significant given the possibility of new mutations against which current vaccines show no or less efficacy. Under the given circumstances, (Moore etal., 2021) show for the case of the UK that vaccination alone is insufficient to contain the pandemic, and that its effect is strongly contingent upon the precise vaccine properties and population uptake. Concluding remarks An essential element of the policy measures to curb the COVID-19 virus is contact prevention. Lockdown strategies to combat the pandemic have a substantial impact on social and economic life and involve immense economic costs. Tools such as testing and vaccination strategies can allow societies more liberties by enabling restrictions to be relaxed in public and private. Our study examines the effects of these control measures on the reproduction rate of COVID-19. 602 M.Fritz et al. 1 3 We set up a theoretical framework to model the effects of policy measures on the reproduction rate. Based on this, we perform a panel analysis using daily data covering 37 OECD countries from March 1, 2020 to May 20, 2021 to assess the relative effectiveness of testing, vaccinations and non-pharmaceutical government response measures. Our estimates provide empirical evidence on the influence of testing, vaccination and specific non-pharmaceutical interventions on the dynamic of the reproduction rate. The simultaneous and data-driven consideration of the instruments provides a comprehensive picture and helps to understand the effectiveness of these measures in the current fight against the pandemic under consideration of different conditions and strategies within countries. For three of our nine NPIs, we identify an estimated relative reduction in the reproduction rate with workplace and school closures showing the strongest effects. A more differentiated analysis shows that the significant effects can only be identified in the case of complete closures. Restricting internal movement likewise reduces contacts and lowers the reproduction rate. With regard to the timing of introduction of social distancing, the lockdown effect is shown to be strongest immediately, decreases with time and approaches zero over time indicating that the adoption time of NPIs and the adjustment to current conditions is important. An early response is critical to counteracting the virus most effectively, as the effect is visible and strongest immediately after introduction. As expected, our results confirm that while vaccinations can mitigate the pandemic, there are significant differences between first-dose vaccinations and full vaccination protection. Further, we can identify diminishing marginal effects suggesting that initially, vaccinations have a strong effect; a certain level of vaccination is required to better control the pandemic. Once this level is reached, additional vaccinations generate only reduced effects. This points to a release of lockdown measures once sufficient vaccines are available and natural herd immunity is achieved. Finally, we find a significant effect for testing. However, a closer look at the interaction with vaccinations shows that this effect only remains significant up to a vaccination rate of about 50%. For as long as vaccines are not available in sufficient quantity or population take up is insufficient, a test, trace, and isolate strategy can help keep the pandemic under control. Our analysis is limited by the type of data utilized. First, the results are based on PCR test statistics. Unfortunately, the number of rapid and self-tests is unknown for the country panel. Therefore, when using the number of PCR tests, we are aware that the effect of testing is relatively strong because the number of rapid and self-tests is much higher than that of the PCR tests. Second, our analysis does not explicitly account for the recent increase in viral variants. The differentiation of variants has only recently been included in the statistics and data is still not widely available across countries. Once this is remedied, future studies could take the differentiation of variants into account and examine whether virus variants change the effectiveness of the instruments. Appendix See Tables6 and 7 603 The effectiveness ofvaccination, testing, andlockdown… 1 3 Table 6 Variables definitions and sources Variable Time Definition Source Reproduction rate Ddaily Reproduction rate Our World in Data COVID-19 dataset, (Arroyo-Marioli etal., 2021) Tests Daily New tests smoothed per thousand Our World in Data COVID-19 dataset, (Hasell etal., 2020) Vaccinations Daily Individuals vaccinated per hundred; individuals fully vaccinated per hundred Our World in Data COVID-19 dataset, (Ritchie etal., 2020) Stringency index Daily Composite measure based on nine response indicators including school closures, workplace closures, and travel bans, rescaled to a value from 0 to 100 (100 = strictest) If policies vary at the subnational level, the index is shown as the response level of the strictest sub-region Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) School closures Daily 0—no measures 1—closures recommended or all schools open with alterations resulting in significant differences compared to nonCOVID-19 operations 2—closures required (only some levels or categories, e.g. only high school or only public schools) 3—closures required at all levels Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) Workplace closures Daily 0—no measures 1—closures recommended (or recommendation to work from home) or all businesses open with alterations, resulting in significant differences compared to non-COVID-19 operation 2—closures required (or work from home) for some sectors or categories of workers 3—closures required (or work from home) for all but essential workplaces (e.g., grocery stores, doctors) Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) Cancelled public events Daily 0—no measures 1—cancellation recommended 2—cancellation required Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) 604 M.Fritz et al. 1 3 Table 6 (continued) Variable Time Definition Source Restrictions on gatherings Daily 0—no restrictions 1—restrictions on very large gatherings (the limit is above 1000 people) 2—restrictions on gatherings between 101–1000 people 3—restrictions on gatherings between 11–100 people 4—restrictions on gatherings of 10 people or less Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) Public transport closures Daily 0—no measures 1—closures recommended (or significant reduction in frequency /route/means of transport available) 2—closures required (or most citizens banned from using it) Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, Hale, etal. (2021) (Hale etal., 2021) Stay-at-home requirements Daily 0—no measures 1—recommendation not to leave house 2—requirement not to leave house with exceptions for daily exercise, grocery shopping, and “essential” trips 3—requirement not to leave with minimal exceptions (e.g. permission to leave once a week, or only one person may leave at a time, etc.) Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) Restrictions on internal movement Daily 0—no measures 1—recommendation not to travel between regions/cities 2—internal movement restrictions in place Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) International travel controls Daily 0—no restrictions 1—screening arrivals 2—quarantine for arrivals from some or all regions 3—ban on arrivals from some regions 4—ban on all regions or total border closure Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) Public information campaigns Daily 0—no COVID-19 public information campaign 1—public officials urge caution about COVID-19 2—coordinated public information campaign (e.g. across traditional and social media) Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, (Hale etal., 2021) 605 The effectiveness ofvaccination, testing, andlockdown… 1 3 Funding Open Access funding enabled and organized by Projekt DEAL. Declarations Conflict of interest There is no funding for this paper and no conflict of interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. References Alfano, V., & Ercolano, S. (2020). The efficacy of lockdown against COVID-19: A cross-country panel analysis. Applied Health Economics and Health Policy, 18, 509–517. Arroyo-Marioli, F., Bullano, F., Kucinskas, S., & Rondón-Moreno, C. (2021). Tracking R of COVID-19: A new real-time estimation using the Kalman filter. PLoS ONE, 16(1), e0244474. https:// doi. org/ 10. 1371/ journ al. pone. 02444 74 Ashraf, B. N. (2020). Economic impact of government interventions during the COVID-19 pandemic: International evidence from financial markets. Journal of Behavioral and Experimental Finance, 27, 1–7. Askitas, N., Tatsiramos, K., Verheyden, B. (2020). Lockdown strategies, mobility patterns and Covid-19. arXiv preprint arXiv: 2006. 00531. Table 7 OLS and System GMM estimations t statistics in parentheses *p < 0.10, **p < 0.05, ***p < 0.01 (1) OLS ΔRR (2) FE ΔRR (3) SYS-GMM ΔRR Tests t-1 − 0.00005** (− 2.089) − 0.00041*** (− 6.180) − 0.00016*** (− 9.278) Vaccinations t-7 − 0.00012*** (− 5.892) − 0.00016*** (− 5.435) − 0.00023*** (− 11.326) Vaccinations (full) t-7 (Vaccinations t-7)2 (Vaccinations (full) t-7)2 Stringency index t-1 − 0.00008*** (− 3.895) − 0.00030*** (− 7.679) − 0.00038*** (− 13.956) R20.014 R2 within 0.028 AR(2) test 1.80 (0.071) Hansen test 36.11 (0.325) Instruments 37 Countries 37 37 37 Obs 3810 3810 3810 606 M.Fritz et al. 1 3 Askitas, N., Tatsiramos, K., & Verheyden, B. (2021). Estimating worldwide effects of non-pharmaceutical interventions on COVID-19 incidence and population mobility patterns using a multiple-event study. Scientific Reports, 11(1), 1–13. Avery, C., Bossert, W., Clark, A., Ellison, G., & Fisher Ellison, S. (2020). Economist’s guide to epidemiology models of infectious disease. Journal of Economic Perspectives., 34(4), 79–104. Baldwin, R. (2020). COVID-19 testing for testing times: Fostering economic recovery and preparing for the second wave. VoxEU. March 26. Available: https:// voxeu. org/ artic le/ testi ngtesti ngtimes. Banholzer, N., van Weenen, E., Kratzwald, B., Seeliger, A., Tschernutter, D., Bottrighi, P., Cenedese, A., Salles, J. P., Vach, W., Feuerriegel S. (2020). Estimating the impact of non-pharmaceutical interventions on documented infections with COVID-19: A cross-country analysis. medRxiv. 2020.04.16.20062141; doi: https:// doi. org/ 10. 1101/ 2020. 04. 16. 20062 141. Bernal, J. L., Andrews, N., Gower, C., Robertson, C., Stowe, J., Tessier, E., Simmons, R., Cottrell, S., Roberts, R., O’Doherty, M., Brown, K., Cameron, C., Stockton, D., McMenamin, J., & Ramsay, M. (2021). Effectiveness of the Pfizer-BioNTech and Oxford-AstraZeneca vaccines on covid-19 related symptoms, hospital admissions, and mortality in older adults in England: Test negative case-control study. BMJ, 2021(373), n1088. Cherif, R., Hasanov, F. (2020). A TIP against the COVID-19 Pandemic, IMF Working Paper WP/20/114, International Monetary Fund: Washington DC Flaxman, S., Mishra, S., Gandy, A., Unwin, H. J. T., Mellan, T. A., Coupland, H., Whittaker, C., Zhu, H., Berah, T., Eaton, J. W., Monod, M., Ghani, A. C., Donnelly, C. A., Riley, S., Vollmer, M. A. C., Ferguson, N. M., Okell, L. C., & Bhatt, S. (2020). Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe. Nature, 584(7820), 25–261. Fontanet, A., & Cauchemez, S. (2020). COVID-19 herd immunity: Where are we? Nature Reviews Immunology, 20(10), 583–584. Gabler, J., Raabe, T., Röhrl, K., von Gaudecker, H. M. (2021). The effectiveness of strategies to contain sars-cov-2: Testing, vaccinations, and npis. arXiv preprint Ge, Y., Zhang, W., Liu, H., Ruktanonchai, C. W., Hu, M., Wu, X.,Song, Y., Ruktanonchai, N. W., Yan, W., Feng, L., Li, Z., Yang, W., Liu, M., Tatem, A. J., Lai, S. (2021). Effects of worldwide interventions and vaccination on COVID-19 between waves and countries, medRxiv. 2021.03.31.21254702; doi: https:// doi. org/ 10. 1101/ 2021. 03. 31. 21254 702. Goolsbee, A., & Syverson, C. (2021). Fear, lockdown, and diversion: Comparing drivers of pandemic economic decline 2020. Journal of Public Economics, 193, 104311. Hale, T., Angrist, N., Goldszmidt, R., Kira, B., Petherick, A., Phillips, T., Webster, S., Cameron-Blake, E., Hallas, L., Majumdar, S., & Tatlow, H. (2021). A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker). Nature Human Behaviour, 5(1), 529–538. Hartl, T., Wälde, K., & Weber, E. (2020). Measuring the impact of the German public shutdown on the spread of Covid-19. Covid Economics, 1, 25–32. Hasell, J., Mathieu, E., Beltekian, D., Macdonald, B., Giattino, C., Ortiz-Ospina, E., Roser, M., & Ritchie, H. (2020). A cross-country database of COVID-19 testing. Scientific Data, 7(1), 345. Jarvis, C. I., Van Zandvoort, K., Gimma, A., Prem, K., Klepac, P., Rubin, G. J., & Edmunds, W. J. (2020). Quantifying the impact of physical distance measures on the transmission of COVID-19 in the UK. BMC Medicine, 18, 1–10. Ke, R., Romero-Severson, E., Sanche, S., & Hengartner, N. (2021). Estimating the reproductive number R0 of SARS-CoV-2 in the United States and eight European countries and implications for vaccination. Journal of Theoretical Biology, 517, 1–10. Kerr, C. C., Mistry, D., Stuart, R. M., Rosenfeld, K., Hart, G. R., Núñez, R. C., Selvaraj, P., Cohen, J. A., Abeysuriya, R. G., George, L., Hagedorn, B., Jastrzębski, M., Fagalde, M., Duchin, J., Famulare, M., & Klein, D. J. (2021). Controlling COVID-19 via test-trace-quarantine. Nature Communications, 12(1), 2993. https:// doi. org/ 10. 1101/ 2020. 07. 15. 20154 765 Lauer, S. A., Grantz, K. H., Bi, Q., Jones, F. K., Zheng, Q., Meredith, H. R., Azman, A. S., Reich, N. G., & Lessler, J. (2020). “The incubation period of coronavirus disease 2019 (COVID-19) From publicly reported confirmed cases: Estimation and application. Annals of Internal Medicine. https:// doi. org/ 10. 7326/ M200504 Linton, N. M., Kobayashi, T., Yang, Y., Hayashi, K., Akhmetzhanov, A. R., Jung, S. M., Yuan, B., Kinoshita, R., & Nishiura, H. (2020). Incubation period and other epidemiological characteristics of 2019 novel Coronavirus infections with right truncation: A statistical analysis of publicly available case data. Journal of Clinical Medicine, 9(2), 538. Mercer, T. R., & Salit, M. (2021). Testing at scale during the COVID-19 pandemic. Nature Reviews Genetics, 22(7), 415–426. 607 The effectiveness ofvaccination, testing, andlockdown… 1 3 Moore, S., Hill, E. M., Tildesley, M. J., Dyson, L., & Keeling, M. J. (2021). Vaccination and non-pharmaceutical interventions for COVID-19: A mathematical modelling study. The Lancet Infectious Diseases, 21(6), 793–802. Nishiura, H., & Chowell, G., etal. (2009). The effective reproduction number as a prelude to statistical estimation of time-dependent epidemic trends. In G. Chowelletal (Ed.), Mathematical and statistical estimation approaches in epidemiology (pp. 103–121). Dordrecht: Springer. Noland, R. B. (2021). Mobility and the effective reproduction rate of COVID-19. Journal of Transport & Health, 20, 101016. Pavelka, M., van Zandvoort, K., Abbott, S., Sherratt, K., Majdan, M., Jarčuška, P., Krajčí, M., Flasche, S. Funk, S., CMMID COVID-19 working group. (2020). The effectiveness of population-wide, rapid antigen test based screening in reducing SARS-CoV-2 infection prevalence in Slovakia. medRxiv 2020.12.02.20240648; doi: https:// doi. org/ 10. 1101/ 2020. 12. 02. 20240 648 Rella, S. A., Kulikova, Y. A., Dermitzakis, E. T., & Kondrashov, F. A. (2021). Rates of SARS-CoV-2 transmission and vaccination impact the fate of vaccine-resistant strains. Scientific Report, 11, 15729. https:// doi. org/ 10. 1038/ s4159802195025-3 Ritchie, H., Ortiz-Ospina, E., Beltekian D., Mathieu E., Hasell J., Macdonald B., Giattino C., Appel C., Rodés-Guirao, L., Roser, M. (2020). Coronavirus Pandemic (COVID-19). Published online at OurWorldInData.org. Retrieved from: ’https:// ourwo rldin data. org/ coron avirus’ [Online Resource] Rodríguez, J., Paton, M., & Acuna, J. M. (2021). COVID-19 vaccination rate and protection attitudes can determine the best prioritisation strategy to reduce fatalities. medRxiv, 2020–10. Sharma, M., Mindermann, S., Rogers-Smith, C., Leech, G., Snodin, B., Ahuja, J., Sandbrink, J. B., Teperowski Monrad, J., Altman, G., Dhaliwal, G., Finnveden, L., Norman, A. J., Oehm, S. B., Sandkühler, J. F., Mellan, T., Kulveit, J., Chindelevitch, L., Flaxman, S., Gal, Y., Mishra, S., Brauner, J. M., Bhatt, S. (2021). Understanding the effectiveness of government interventions in Europe’s second wave of COVID-19. medRxiv. 2021.03.25.21254330; doi: https:// doi. org/ 10. 1101/ 2021. 03. 25. 21254 33 Taipale, J., Romer, P., Linnarsson, S. (2020). Population-scale testing can suppress the spread of COVID19. medRxiv. 2020.04.27.20078329; doi: https:// doi. org/ 10. 1101/ 2020. 04. 27. 20078 329 Thompson, M. G., Burgess, J. L., Naleway, A. L., Tyner, H. L., Yoon, S. K., Meece, J., Olsho, L. E. W., Caban-Martinez, A. J., Fowlkes, A., Lutrick, K., Kuntz, J. L., Dunnigan, K., Odean, M. J., Hegmann, K. T., Stefanski, E., Edwards, L. J., Schaefer-Solle, N., Grant, L., Ellingson, K., … Gaglani, M. (2021). Interim estimates of vaccine effectiveness of BNT162b2 and mRNA-1273 COVID-19 vaccines in preventing SARS-CoV-2 infection among health care personnel, first responders, and other essential and frontline workers: Eight US locations, December 2020–March 2021. Morbidity and Mortality Weekly Report, 70(13), 495. Vannoni, M., McKee, M., Semenza, J. C., Bonell, C., & Stuckler, D. (2020). Using volunteered geographic information to assess mobility in the early phases of the COVID-19 pandemic: A cross-city time series analysis of 41 cities in 22 countries from March 2nd to 26th 2020. Globalization and Health, 16(1), 1–9. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.