Purpose of the analysis
The recent rise in the number of COVID-19 infections, which was most likely caused by new virus strains, stirred heated public debates about the need to enforce tougher restrictions in order to prevent the virus from spreading. In this context, as a result of the recent meeting of the Supreme Security Council, President Sandu suggested to impose a 2-week state of emergency and for the Government to develop the list of restrictions regarding social and economic activities, which would be then submitted to the Parliament for consideration and approval. The idea of establishing these restrictions (or a lockdown) is controversial per se, as its impact will be felt by everyone, implying costs for some and benefits for others. On the one hand it could lead to economic and social costs, with particular damage to small enterprises and employees with lower salaries, especially if we take into account the population’s low income and the economy that is barely recovering from the most severe recession in the last 20 years, while on the other hand a lockdown could relieve some of the pressure on the health system, prevent infection spreading and allow saving people’s lives. Nevertheless, the heated arguments about this issue miss a cold review of the effects caused by the lockdown of 17 March - 16 May 2020, which could be used as a yardstick for understanding the effects of a potential lockdown in 2021. This analysis estimates those effects on the level of infection cases and deaths caused by COVID-19, as well as on the major economic indicators, by means of precise econometrics and statistical techniques.
Impact of the lockdown on the health situation
Methodological set up
In order to estimate the lockdown impact on the health situation, we specify two very simple econometric models, which proved to be very precise. The models allow to estimate the key drivers of infection and death rates caused by Covid (please refer to Annex A for more detailed specification of econometric models). Respectively, these are:
- External factor. Taking into account that the pandemics is an external shock for the country, it is necessary to include into the model a variable that controls for the regional and world tendencies of virus spreading, because they affect the situation in Moldova. The most relevant indicator in this regard is the infections' and death rates registered in Romania, taking into account the geographic and economic proximity of both countries. Thus, the model takes into account the epidemiological situation in Romania as an explanatory factor for the epidemiological situation in Moldova. Inclusion of this variable excludes the necessity to add variables for other countries because, given the global nature of the pandemics, there is a high correlation among countries. In such conditions, in order to ensure a proper fit of the model, it is sufficient to include one single variable, which has the most direct influence on the situation in Moldova.
- Inertial factor. Taking into account the nature of the virus spreading, the number of infections and deaths caused by Covid also depends on the past tendencies (e.g. the higher the number yesterday, it is more likely that today and tomorrow the numbers will also be higher).
- The lockdown. The model will take into account the impact of the lockdown during the state of emergency imposed during March 17 – May 16, as it is assumed that it allowed to slow down the spreading of the virus. In this way, we will be able to estimate the efficiency and impact of that lockdown, which could serve a valuable input into the ongoing debates about the potential lockdown in 2021.
- Post-lockdown period. It is assumed that after lifting the restrictions on May 16, the virus started to spread at a higher speed, leading to a spike in number of infections and deaths. Therefore, we include into the model another binary explanatory factor that controls the period of one month past two weeks from the lockdown being lifted (after the restrictions were lifted and past the virus incubation period, we had a rapid increase in the number of infections and deaths). This will allow us to estimate whether the lockdown was lifted at the right moment or it was too early, which will also serve as an important lesson learned in 2020 for the potential lockdown in 2021.
- Weekdays. Taking into account that on Sundays the number of reported infections and deaths tends to be lower, which is compensated by higher numbers on Mondays, we include the weekdays in order to control for this seasonal factor and increase the accuracy of the model.
The estimation results reveal that the 2020 lockdown allowed decreasing the number of new infection cases by 20% and the number of deaths by an impressive 45% (regression results are presented in Annex B, while the forecasts based on this model are presented in Annex C). If we extrapolate these figures to the situation in 2021 and take into account the forecasts on infections and deaths for the upcoming 30 days, we can estimate that imposing a lockdown would decrease the number of new cases by around 12.3 thousand and, most importantly, it will allow saving around 287 people. These figures apply strictly to the following 30 days, during which we forecast that, if a lockdown is not imposed, another 61.5 thousand infection cases and 1,432 deaths would be registered. Moreover, the lockdown impact on decreasing virus spreading and reducing the number of deaths will be felt over the following months, because decreasing the number of infection cases by around 12.3 thousand during these 30 days will protect other 36.9 thousand people that potentially could be infected in chain (based on the hypothesis that one infected person would infect three other people). Given that the share of deaths out of total cases, according to the forecast for the next 30 days is around 2.3%, we can conclude that additional 849 deaths will be avoided. All in all, imposing a 30-day lockdown will allow avoiding the infection of around 49.2 thousand people and saving 1,136 lives.
Impact of the lockdown on the economic situation
Estimation of the 2020 lockdown impact on the economic situation is trickier and we will use a slightly different approach. The reason is that the economic dynamics in 2020 was obscured by the overlapping of the pandemic crisis and two other crises: a slowdown in the economic growth noticed since the autumn of 2019 and the drought of 2020 that substantially the agricultural production and agro-industrial exports. Respectively, in order to estimate the net impact of the lockdown it is necessary to ‘control’ all other factors that also influenced (mostly negatively) the economy in 2020. Respectively, in order to estimate the net impact, we will compare the de facto developments during the lockdown period (17 March – 16 May 2020) with the potential evolutions that would have happened if there wasn’t any lockdown. Also, we will analyze some key macroeconomic indicators that are available on monthly frequency, in order to analyze to what extent their dynamics was affected during the lockdown period (17 March – 16 May 2020).
First of all, we will estimate the lockdown impact on budget receipts. For this purpose, we will analyse the impact over the revenues from taxes (mainly VAT, excise duties and income tax), which are proportionate to the level of consumption and overall economic situation; we will also analyse the impact on the collection of mandatory social and medical insurance contributions, which are proportionate to the level of salaries and employment rate in the economy. The estimates confirm a negative shock, but not a dramatic one, of the lockdown on both sources of revenue. Thus, in the lockdown months (March – May 2020) the collected taxes were MDL 1.3 billion lower than the estimated level if no lockdown would have been imposed3 . When it comes to mandatory social and medical insurance contributions, the collected amount was MDL 373 million lower than the amount that would have been collected in the absence of a lockdown. Thus, we can state that the net impact of the lockdown on the national public budget is estimated at around MDL 1.7 billion (or about 3% of total public revenues) – calculated as the difference between actual collection during March – May 2020 and what could have been collected if no lockdown was imposed. At the same time, we do not see a dramatic effect and the dynamics in the following months showed a marginal improvement of budget receipts.
The next indicator is the Industrial Production Index (IPI), which is closely related to the main macroeconomic indicator, which is unfortunately available only with quarterly frequency – the Gross Domestic Product. During the lockdown months this index registered a steep decline, which was much more dramatic than the estimated development in case of a no-lockdown scenario. The biggest blow happened in April 2020, when the volume of industrial production decreased by 26% compared to April 2019, while in a no-lockdown scenario the decrease could have been 2.5 times smaller (however, the index would have declined anyway under the overall impact of the pandemic). At the same time, it is worth mentioning that the industrial activity started to recover slowly immediately after the lockdown was lifted, which revealed a certain adaptability and flexibility of the industrial complex. Thus, despite the dramatic repercussions of the corona crisis, decreased internal and external demand, but also the lack of actual support of the Government for companies and entrepreneurs4 , the industrial activity finished year 2020 close to zero (in December the index decreased by only 1.8% and in November 2020 it even registered a 2.8% increase compared to the similar period of the previous year).
Another important indicator that highlights both the level of consumption and economic activity is the import dynamics. When it comes to this indicator we also notice a sudden decrease in April 2020 (-45%) and May 2020 (-32%) compared to the previous year, which was almost twice as deeper as it would have been in a nolockdown scenario. In this respect, it is necessary to mention that the descending trend started way before the lockdown. In case of this indicator we also notice a visible recovery in the post-lockdown period. Similar dynamics (V-shaped) are also typical for other indicators that measure the level of economic activity in the country (e.g. domestic trade in goods and services), which reveals that most companies perceived the lockdown as a temporary measure and after it was lifted they resumed their activity (although only partially), which led to a certain compensatory growth of the economic activity (the growth is insufficient to overcome the corona crisis, but a marginal recovery compared to the lockdown period is visible).
A new lockdown: to be or not to be?
The econometric estimates confirm the beneficial impact of the 2020 lockdown on the health situation. It allowed decreasing the number of new COVID-19 cases by around 20% and reducing the number of deaths by around 45%, compared to a no-lockdown scenario. Surprisingly, its economic costs were not as dramatic as it is often claimed. It is true that 2020 was the year of one of the worst recessions in the last 20 years, but this was caused by the overlapping of two crises: the drought and the COVID-19 impact on the internal and external demand. The lockdown had a certain net effect that had worsened the situation for two months, but this effect was compensated by a marginal economic recovery, during the post-lockdown period. The budgetary impact of the lockdown was also far from dramatic, being estimated to about MDL 1.7 billion (about 3% of total public revenues in 2020), which was also followed by a marginal recovery of budget receipts over the following months. Given the analysis of the 2020 lockdown impact, we can state that a potential lockdown for only 30 days in 2021 will prevent the infection of around 49.2 thousand people and save over 1,100 lives. Taking into account that Government’s priority zero should be to prevent the virus spreading and save human lives, in circumstances where the number of infection cases and deaths grows rapidly, it is obvious that imposing a new lockdown is necessary. At the same time, in order to mitigate the economic costs, which could be higher than in 2020, because companies and population had mostly exhausted their reserves, the lockdown needs to be correlated with certain compensatory measures focused on two goals: (i) access to financing for companies, by issuing state guarantees; and (ii) compensation of salary losses for employees, to an extent of around 75%.