Closed: Crises, politics and statistics (Closed)

Official Statistics COVID-19

This special discussion on the SJIAOS discussion platform focuses on the roles of Official Statistics in the context of the COVID-19 crisis. It states the important role that official statistics will have to play in the phase of world-wide recovery from the pandemic and the rapid investments and actions that are needed to fulfill properly this role. It also emphasizes the importance of investing in achieving the objectives of the SDG indicators, the need to develop new statistics and using modern data sources, and last but not least the establishment of a new role of official statistics in the public statistical infrastructure.

Official Statistics in the context of the COVID-19 crisis

This discussion on the SJIAOS discussion platform (www.officialstatistics.com) is based on the following statement[1].

Now epidemiology is most important, in the next phase other statistics will become relevant again!

The enormous importance of statistics in the preparation of difficult political decisions has seldom been as clear as it is now in the current phase of the pandemic crisis.

While in the current, first phase of crisis management, where essentially the health risks of this epidemic vis-à-vis the absorption capacities of national health systems were sufficient to justify the measures taken, this will be completely different in the next phases. In order to decide when and how quarantine and isolation can and should be ended and transferred to other forms, important other aspects will have to be considered, such as domestic violence, unemployment, collateral illnesses (depression, suicide), economic consequences (despite or because of the financial support measures), ecological facilitations. At the same time, all the topics that were given lower priority in the short-term crisis mode will be on the agenda again. Achieving the UN Sustainable Development Goals, overcoming poverty and inequality, protecting the climate and biodiversity and other goals set out in the global strategy until 2030 will not be easier to achieve after the COVID-19 crisis, quite the contrary. We (politics, societies at large) will be faced over the coming years with multiple and extremely complex decisions that require an unprecedented level of speed, quality and reliability of new metrics in more or less all fields: economy, distributive justice, unemployment, agriculture, transport, energy, health, education, to name but a few.

Not just the current, but many new statistics are needed

The fallout from the current crisis will be found in all areas and variables observed by statistics. As if that wasn't enough; there will be an urgent need for new statistics, more urgent than ever before. New statistics in this case also means new services, new forms of communication, new software, new networks, new forms of statistical education and last but not least: new data sources (i.e. new surveys and other means for observing phenomena of interest).

SDG’s information as the all-encompassing framework

Now would be the moment to take the UN SDGs seriously and interpret them as what they should be and could achieve. They could enable a great transformation (in terms of change of behaviour and long term societal development). However, first of all, SDG-relevant information (indicators, accounts, statistics) needs to be seen as the all-encompassing and defining methodological framework into which all statistics should fit. Secondly, the existing statistical information would now need to be rapidly supplemented with very short-term solutions and, in part, inputs from external sources. Thirdly, open data, publications and evidence-based networks and collaborations should be actively promoted. Fourthly, the Codes of Practice should urgently be adapted to this new and considerably modified situation (for example, also by means of appropriate information and communication of the quality profiles, i.e. labelling). All in all, official statistics should actively contribute to mastering the current short-term crisis and the development of a SDG compatible exit strategy, which is directly linked to the fact that the SDGs (all of them) are regularly monitored by statistics.

Official statistics should be on board to help in crisis management

In the near future, politicians will be confronted with extremely complex decision problems. The technical consulting teams for crisis management will therefore be multidisciplinary. It would be of major importance for statisticians to be on board here to contribute their various potential strengths, such as wealth of methods, experience in various applications, comprehensive information infrastructure, long time series, solidity and trust.

A new role for official statistics in the public statistical infrastructure

All of this will require innovation and an openness to solutions that will not necessarily fit into the forms and frameworks established in quieter times. And all this presupposes that the statistical institutions are set up in such a way that they can meet these new requirements. This would require a public statistical infrastructure that is at the same time professionally independent, high-performing and innovative.

‘Official statistics’ has to adapt quickly to the changing circumstances

What will be the reservations about such a bold strategic positioning? A characteristic argument will be linked to the question of what official statistics actually is and how far one can diversify without endangering the brand core. For this discussion it is crucial to remember official statistics’ DNA[2] and that there is no such thing as the quasi fateful definition of what official statistics means. The status we know today is the result of an evolution over many years, decades, in terms of institutions (who), programmes (what) and methods (how). This development was and is influenced by political events, scientific progress and new data sources. Most of the time, this evolution is very slow and steady, so we are hardly aware of it. Nevertheless, it does happen, which we can make clear to ourselves, for example, by the fact that a statistician before the economic crisis of the 1930s and before Keynes could hardly have imagined that we now have GDP as an important indicator in our portfolio. Alain Desrosières has worked out in detail the interrelation between the state and statistics; with him one could even speak of statistics for ‘the making of the state’.

Statistics will play an active role in shaping the world of tomorrow

What the world will look like after the current crisis and whether or not we move towards sustainable development (or to life as it was) is completely open today. What is clear, however, is the need for statistics to play an active role in this making of the world of tomorrow. In doing so, it is important to open up widely to new possibilities and necessities. Is it possible to create new forms of cooperation with other players, located outside the traditional city walls, which could benefit from the know-how or the reputation of statistics (e.g. by assigning appropriate labels, such as "operated by Official Statistics")? Should consideration be given to strengthening the service branch of the official statistics industry (e.g. by providing software, awarding quality certificates, acting as data steward or helping to set up and conduct statistical surveys)? Could it for example be a task for official statistics to establish a reporting system for biodiversity comparable to the system of population statistics?

Rapid action and substantial investment are needed

All these innovations presuppose that one has the required capacities to launch and realize them. After decades of austerity policy, however, statistical institutes hardly have the reserves and the necessary research and development capacities that are so urgently needed right now. In the health care system, the comparable deficiency has now entered the public debate, although this cannot yet be observed with regard to similar problems in statistics. Taken together, it can be concluded that a call for rapid action must be made right now. It must not take months and years for coordination processes. Rather, a wide variety of ideas and initiatives should sprout in a decentralized way and at the same time international networks and cooperation should be promoted. Above all, however, official statistics must draw the attention of the public to their importance and their miserable situation, so that an appropriate share of the enormous financial sums currently being spent on crisis management is also invested in improving adequate statistical evidence. 



[1] The text of this statement has been prepared by Walter J. Radermacher and subscribed by the president of the IAOS, John Pullinger.

[2] See W.J. Radermacher „Official Statistics 4.0” https://link.springer.com/book/10.1007%2F978-3-030-31492-7

This discussion is now closed


Comment on:
17/06/2020 20h23

I want to share in this platform that the FAO Statistics Division's Agricultural Censuses Team have published two publications related to the impact of COVID-19:

• National agricultural census operations and COVID-19
• Impact of COVID-19 on national censuses of agriculture (Status overview)

The publications are available at the FAO’s Covid-19 page at http://www.fao.org/2019-ncov/resources/policy-briefs/en/. You can find them once you search the topic = ‘statistical systems’.

The information presented is the result of a rapid appraisal and informal consultations with national agricultural census authorities. This first publication describes the challenges and good practices. The second overview covers 112 countries and shows how much and how far have been affected agricultural census activities in the countries.

Thank you.

Jairo Castano
Leader, Agricultural Censuses Team
Statistics Division (ESS), Economic & Social Dev Dept (ES)
FAO, Rome

M. De Smedt

Comment on:
04/06/2020 19h50

The attached wider article describes how official statistics, in casu European statistics on health and health care can contribute for measuring COVID-19 and its consequences on health status and health care. This pandemic might be an opportunity for countries to learn from each other’s experiences and to seek more common ground for quick, adequate and comparable data (and management) in case of a crisis, in casu an urgent problem of public health. Knowledge, expertise and common methods and tools are available, but we need new solutions, which allow for faster reactions and a quicker response, that might not entirely correspond to all quality requirements, but which are however delivering best possible information with the highest possible doses of quality assured by official statistics.


Measuring the COVID-19 pandemic and its consequences
Official statistics on health and health care in the European Union

The COVID-19 pandemic has raged through populations over the globe from east to west and has put an immediate and important strain on health systems all over. The pandemic concerns us not only at a personal level but also at family and community level. It is a true public health problem at all geographical scales.

This exceptional situation makes governments, media, businesses, NGOs and the public at large more aware of the importance of ‘health and safety’. The need for appropriate surveillance, monitoring, and diagnosis and treatment of potentially fatal - infectious – diseases has never been more at the forefront of the debate.

The objective of this article is to reflect on actions taken at European level, on type of data needed for assessing and managing the COVID-19 pandemic and its consequences, and in particular on how European statistics on health and health care play their part in the data collection.

What actions on public health are taken at European level?

Up to now, the competences of the European Union (EU) in the field of public health are limited. EU Members States (MS) hold primary responsibility for organising and delivering health services and medical care and the related manpower and financing. EU health policy therefore serves to complement national policies, and to ensure health protection in all EU policies.

EU policies and actions in public health1 concern inter alia promoting the health of EU citizens, exchanging information, supporting the modernisation of health infrastructure, and supporting countries in tackling common challenges such as in the case of pandemics. For the latter, in case of emerging health threats - in casu infectious diseases - the European Centre of Disease Control (ECDC)2 has an important coordination role to play in the field of monitoring and rapid response.

But organising and delivering health care services remains the responsibility of the national governments. This explains partly the different approaches and measures taken by the EU MS in this crisis situation. However, for statistics on health and safety, there is a statistical legal ‘package’ at EU level, committing the EU MS to apply common standards and methods for arriving at reliable and comparable European statistics.

1 https://ec.europa.eu/info/departments/health-and-food-safety_en 2 www.ecdc.europa.eu

Rapid response and epidemiology

The immediate management of the crisis required a rapid response from various actors: virologists, epidemiologists, nurses, doctors and politicians. The health care systems and personnel have to provide the necessary diagnosis and treatment; epidemiologists and virologists have to monitor the situation in the population and give expert advice; politicians are expected to decide on crisis management; and last but not least the public has to follow the rules for social distancing, hygiene etc.

As for the immediate assessment and monitoring of the impact of the SARS-Cov-2 virus on public health, first data are provided by hospitals. Since the beginning of the crisis, key indicators that describe the evolution of the pandemic are: number of tests carried out, number of persons tested positive and/or for this reason admitted to the hospital, number of patients in Intensive Care Units (ICUs) number of deaths related to COVID-19 and number of patients having left the hospital. Very often the figures on these key indicators are published on a daily basis and they dominate the news bulletins all over the world.

Because of the urgency, countries initially relied on rapidly available data, which do not necessarily entirely fulfil the quality requirements of official statistics. However, these data were undeniably useful in the early days of national crisis assessment and management, and provided an acceptable basis for assessing the development of a pandemic in a given territory on a daily basis.

However, these daily ‘national’ figures were (and still are) in most cases not covering the whole population, nor were they internationally comparable and this still is the situation today. The number of infections depends on the testing strategies used and these might differ between counties. And in interpreting the figures between countries, other factors such as population density have to be taken into account.

For a number of countries, the respective figures (infections, deaths, ...) for nursing and residential care facilities (such as homes for the elderly), and in private households are not, or only partly, included in the ‘national’ data. In addition, the figures on deaths might include as well deaths due to ‘COVID-19’ as deaths with ‘COVID-19’.

We should also bear in mind that, in this crisis situation, Goodhart’s law might apply: “"When a measure becomes a target3, it ceases to be a good measure."

The quick response in this health crisis required not only figures but also much more information, such as quick results of relevant studies and research on the

3 to be understood in this crisis as the lowest possible numbers for infections or deaths.

virus, availability of hospital beds and beds in Intensive Care Units (ICU) in particular and of respirators, results of modelling, lessons learned from earlier outbreaks (i.e. in Taiwan, South Korea), expert opinions on possible medications and on plans on preparing vaccines.

Gathering this type of information analysing it for risk assessment and using it for preparing adequate risk management is the field of epidemiology. This requires efficient cooperation between different providers i.e. universities, ministries of health, national statistical institutes and national scientific institutes on public health).

The role of science and of scientific experts is rediscovered and regaining its importance. ‘We rely on scientific expertise’, is now like a ‘mantra’ in many of the governments’ press releases and also in many daily conversations.

The role of official statistics

Providing high quality figures representative for the relevant geographical level (European, national, local) concerned is the field of official statistics.4 For some of the above-mentioned key indicators, official – national and European - statistics exist, such as on hospital discharges, health status, mortality and causes of death.

European statistics build on commonly agreed principles as laid down by European legislation5 and in the “European Statistics Code of Practice (CoP)”.6 The establishment of official European statistics, is organised in the European Statistical System (ESS) which involves both Eurostat7, the National Statistical Institutes (NSIs) and Other National Autorities (ONAs)8 responsible in each MS for the development, production and dissemination of European statistics.

For the different statistical fields European Regulations (decided by Council and European Parliament) set out the scope and type of European statistics to be provided as well as the time and frequency and methodology to be used. MS also need to provide regular quality reports. The details of the statistical data collections are further set out by Commission Regulations. The Regulations are an important step forward to obtain high quality statistics that are representative for and comparable between EU MS and over time.

4 Radermacher W.J. (2020) Official Statistics – An introduction in Official Statistics 4.0. Springer, Cham
5 European Regulation 223/2009 on European statistics
6 https://ec.europa.eu/eurostat/web/products-catalogues/-/KS-02-18-142
7 the statistical office of the EU
8 for health statistics, the ONAs are mainly the Ministries of health and the National Institutes for Public Health.

Since 2008, an EU framework Regulation9 has laid down the legal requirements for
European statistics in the field of public health and health and safety at work10 According to this EU law on health statistics, MS have to send their official statistics inter alia on causes of death, health status and life styles and on health care facilities and the use of it, on health expenditure and financing, to Eurostat. For each of these specific health domains, the details are laid down in Commission Regulations.

From the viewpoint of business architecture, European health statistics are mainly based on administrative data (death registrations, hospital data..), which have the advantage of good coverage, but these registers are often rigid and cannot easily be adapted to new information needs. An important other source for health statistics are population surveys: data are obtained on a representative sample of persons and the variables to be questioned could be adapted according to the needs of the situation.

For most of the European health statistics, the ESS has set up its data collection in cooperation with international organisations as WHO and the Organisation for Economic Cooperation and Development (OECD).

How could official statistics help in counting patients and deaths?

Let us first look at the official causes of death statistics. The official procedure – on the basis of the World Health Organisation (WHO)’s rules using the International Classification of Diseases (ICD)11 - is as follows: when the responsible physician notifies the death of a person, he/she notes (or at least is required to note) the sequence of the events that have led to the death of the person, on the official death certificate.

This certificate is then sent to the designated national authority (often after having first been passed to the local authority). It is at this authority that a specialist in coding12 applies the international coding rules established by WHO to select the underlying cause of death (UCD), which - according to general medical standards – is accepted as ‘the’ main cause of death. Other diseases might be ‘contributory causes’ but not the main cause.

This is an internationally agreed procedure, which – for EU MS – is laid down in a specific Commission Regulation13 that prescribes the timing, frequency, scope and

9 and the corresponding Commission Regulations
11 includes now an international emergency code for COVID-19 disease 12 some countries use automated coding systems
13 http://data.europa.eu/eli/reg/2011/328/oj

variables of the data which countries should provide. For checking statistical data quality and completeness, a standard data validation tool14 should be followed, The objective of this procedure is to have quality assured comparable on causes of data death over time and between all MS.

But this whole procedure is lengthy and cumbersome, because it is, in most countries, part of a civil registration system required for assuring completeness and quality of population registers. This is why, for this crisis, hospitals are immediately reporting the COVID-19 deaths occurring in their establishments to (their national) authorities. This is the major part of the “fast” COVID-19 death reporting,” however hampered by quality problems.

Some countries also include the COVID-19 deaths reported directly by General Practitioners (GPs) and/or reported by nursing and residential care facilities. Not adding these additional (out-of-hospital) deaths leads to an under-reporting of COVID-19 deaths. But also over-reporting might be the case, such as in nursing homes where all deaths are counted without specifying whether these patients were tested COVID-19 positive or not.

Several national (and regional) statistical authorities (FR, UK, IT, NL) have stepped up efforts by providing, by means of their existing official registration system, ”fast” official statistics on COVID-19 deaths in order to amend and complete the early figures.

Time series analysis of overall mortality and in particular, comparing official statistics on ‘general average mortality’ (calculated over the last years) by month week or day, and for different age groups and sub-national entities, with the ‘current mortality figures’, could give an indication of the possible excess-mortality at population level.
In the COVID crisis, this excess mortality could partly be attributed to deaths linked to COVID-19 (and its complications)15.

So, in the peak of the crisis, excess mortality gave a quick and inter-country/regional comparable approximation of the severity of the pandemic. But it is only when all official causes of death statistics will be available for all countries that a reliable inter-country comparability on COVID-19 deaths will be possible.

Besides COVID-19 deaths, hospitals also give the number of persons hospitalised (daily admissions and cumulative number), the number of persons in ICU and the number of people having left the hospital (daily and cumulative number). Again, the regular routine reporting of official statistical series for hospital data, had to be stepped up to a daily direct reporting from the hospitals to the authorities.

14 https://ec.europa.eu/eurostat/cache/metadata/en/hlth_cdeath_esms.htm
15 Part of this extra mortality might also be due to delayed health care for other than COVID- 19 related deaths.

As for the number of persons tested positive in a given country, this figure depends on total tests carried out and on testing strategies. While in the initial phase of the crisis, only persons with severe COVID-19 symptoms were tested, later on, in the exit strategies, persons with even vague COVID-19 symptoms are also tested (and their contact traced and tested when needed). Testing strategies vary between countries and over time and this hampers national representativeness, reliability of time series and international comparability of data on number of infections.

Nevertheless, the figures on positive tests (infections) are used for another important indicator to monitor the pandemic and the exit strategies: the ‘reproduction rate R’. This number indicates how many persons - at a given time – are newly infected by an infected person. Some countries (D, A, DK) communicate this R at regular intervals. In general the R can be calculated by dividing the number of new infections of the last four16 days by the number of new infections of the four days before. But again, differences in testing strategies hamper inter-country comparability.

No legal basis exists yet for official statistics on morbidity, but a common method is already available with the purpose of obtaining data on incidence and prevalence of diseases across the EU. For infectious diseases the data are directly available through the ECDC.

Countries are now carrying out their exit strategies, out of the lockdown situation. An important step will be when cross-border travel will be allowed again. The idea of opening borders in a first step between countries with similar infection rates has been raised. But for this, good inter-country comparable figures for the key health indicators in this pandemic (infections, R, deaths) are even more needed.

And what about asking people directly about their health and personal life ?

Through interviewing a representative sample of a given population, information could be obtained about the characteristics of that population by means of the systematic use of statistical methodology. The ESS has established a number of population surveys17, which provide a wealth of data on commonly agreed socio-economic variables and on domains such as living conditions, labour situation, well-being, education and health and safety.

Health interview surveys (HIS), based on a specific Commission Regulation are the source for European health statistics on various aspects of health status (physical and mental health), life styles (diet, alcohol consumption, physical activity) and use

16 four days is the average time for an infected person to infect someone else. 17 Based on EU Regulations, i.e. Labour force survey, EU-SILC, HIS

Through interviewing a representative sample of a given population, information could be obtained about the characteristics of that population by means of the systematic use of statistical methodology. The ESS has established a number of

of health care services. Two waves of European HIS have been carried out and the third wave is ongoing.

Setting up a HIS at EU level is a lengthy procedure, requiring a lot of preparation, and procedures to be set up, but – once done – it offers – in addition to the data - general statistical expertise on methodologies, surveying design, common variables and questions (in all EU languages), reporting mechanisms and quality assurance procedures18.

Some MS have used this common survey expertise to set up fast ad-hoc surveys on COVID-19 (launched or about to be launched in BE, ES, UK and FIN). The objective of these special ad-hoc surveys (sometimes combined with testing and taking of blood samples) is to provide specific and reliable COVID-19 health data – at population level - mainly on the occurrence of symptoms, and infections, but also on general health status and well-being of the population in this crisis and on the impact of the virus and of confinement measures on people’s lives.

Health care at the spot

In the last decades, the sector of social services in general and health care services in particular has suffered from insufficient structural and financial resources. And now suddenly it has become a matter of the utmost importance. Health and health care are now at the centre of the debate and promises are made to examine related health care structures, processes and financing and to remedy situations of understaffing and of important strain on health care workers.

In the domain of health care, official statistics could contribute by delivering data on hospital beds and ICU beds, on number of health care personnel available, and on use of health care services by the population19. For a better understanding of the data on health care systems (and their components: type and kind of services, functions carried out and manpower included), detailed descriptions of these systems in the EU MS are available20 .

These detailed descriptions are a prerequisite for a better monitoring, planning and evaluation of the pressure and the impact of the COVID-19 crisis on the national health care systems. Many health workers ask for a serious evaluation and, where needed, ‘reshaping’ of the whole or part of the health care sector. At a later stage, official statistics will show us the outcome of different approaches followed by governments.

18 https://ec.europa.eu/eurostat/web/products-manuals-and-guidelines/-/KS-… 19 https://ec.europa.eu/eurostat/web/covid-19/population-health-latest-rel…
20 https://ec.europa.eu/eurostat/statistics- explained/index.php/Healthcare_resource_statistics_-_beds https://ec.europa.eu/health/state/country_profiles_en

The COVID-19 pandemic is affecting all sectors of economic and social life
Apart from the ‘medical’ COVID-19 victims there are also the ‘economic’ COVID-19 victims. The virus has taken us in speed, and a quick reaction was needed. Most governments took drastic measures and decided for a (semi) lockdown.

This lockdown has created a ‘ravage’ at an even larger scale: this pandemic has dramatic and far-reaching consequences on many other aspects of our personal lives and on so many sectors of our societies at local, national and global level. For monitoring and managing the broader effects of the crisis, reliable information on so many other sectors is needed.

The good news is that - as mentioned by former Eurostat DG W.Radermacher21 - we already have an internationally agreed guide that could help us out of the pandemic in a way that also has a larger focus on the development of our world.

The United Nations’ 2030 Agenda provides the worldwide framework that shows the way to go for ensuring sustainable development by working towards 17 Sustainable Development Goals (SDGs) and 169 targets. This larger framework will allow us to monitor also other existing or emerging crises such as starvation or climate change.
In the ESS, progress on sustainable development is monitored using a set of 100 indicators22, yearly updated on the basis of the most recent official statistics available. As an integral part of its strategy to implement the United Nation’s 2030 Agenda and the SDGs, the European Commission adopted last year its a communication on the Green Deal23 .

This Green Deal is specified through a list of key actions with specific deadlines and progress will also need to be monitored. Nowadays, European statistics – established by European statistical legislation - are available through Eurostat’s website24 and this for many fields such as economy, agriculture, businesses, trade, environment, population, living conditions, labor, health, education, ICT, tourism, energy and transport. Eurostat also lists ‘Special guidelines and methodologies on COVID-19 support for statisticians’ on its website25.

21 www.officialstatistics.com
22 http://ec.europa.eu/eurostat/web/sdi/indicators
23 COM(2019)640 final of 11.12.2019
24 http://ec.europa.eu/eurostat/web/main/home
25 http://ec.europa.eu/eurostat/data/metadata/covid-19-support-for-statist…

Regular official statistics give reliable and comparable figures on many socio- economic variables, data of top quality and comparability, but not ‘at the spot’ and too slow for providing early warning. However, a previous crisis - the financial crisis of 2008 - has stimulated statistical authorities to develop methods for providing key data on economy, labour and income in a much faster way, such as through estimates, or by publishing provisional data or a combination of both.

So besides the regular routine monitoring, the ESS could now give fast and provisional data on key socio-economic indicators. The consequences of the pandemic are already observed, such as a rising unemployment, closing of businesses, decrease in household income and increasing risk of poverty.

Lessons learned

In monitoring and managing this COVID-19 pandemic we could list a number of lessons learned.

First, health is important. Surveys on wellbeing26 already showed that ‘being in good health’ is in the top three of what people consider as important for their life satisfaction and wellbeing. However, this pandemic goes beyond the personal benefits; good health at personal level now goes hand in hand with the general concern of ‘public health’. We are all concerned by this pandemic: the health of my neighbour, my family, my community is also important for my own health.

Second, good information on public health, in casu this pandemic, requires good cooperation between providers of a vast array of disciplines (statisticians, specialised institutes, universities and administrations). All have to work together to give the best possible ‘expert advice’. European health statistics have their role in providing reliable and comparable figures on health and health care.

Third, this crisis again demonstrates that providers and users of (health) data and information such as politicians have their own final responsibilities with their own separate mandates, but a continued dialogue between them is needed for effective crisis management.

Fourth, as shown by the often daily reporting by countries on key indicators reliable and comparable data are a cornerstone in the information used by politicians and authorities. Official statistics are/ should be readily available on many domains, going from economy and trade, to environment, labour, living conditions and health and safety. In official statistics we have to be prepared: what for some might not have been a priority yesterday, suddenly becomes a priority today. Besides these
26 https://ec.europa.eu/eurostat/en/web/products-statistical-working-paper…- 16-005

routinely provided statistics, NSIs and ONAs should be flexible enough to respond quickly to - possible new - data requests.

Fifth, the COVID-19 crisis has shown that official morbidity statistics, reliable and comparable statistics on the vast majority of diseases at population level, are still missing. For proving these data, the hospitals – and in some cases sentinel practices of GPs - are the main (and often only) source. But a good recording of morbidity data of other parts of the population, such as in the general public and in nursing homes and other residential and care facilities, is missing in many countries. And, when data are available, they are hardly comparable between countries because the organization of social and health care services differs as well as the concepts, definitions and classifications used.

In general, more coordination between countries is needed. For cross-border (health) problems we need cross-border solutions. This pandemic might be an opportunity for countries to learn from each other’s experiences and to seek more common ground for quick, adequate and comparable data (and management) in case of a crisis, in casu an urgent problem of public health.

For instance, exchanging information at EU level, on the processes of data collection on COVID-19 deaths in MS, could be a good start. For each country, the NSI could describe the procedure used. Together with the ONAs involved, NSIs could then work out – in the context of the ESS - a procedure for a fast reporting of deaths in times of a health crisis.

So in general, we need new solutions, which allow for faster reactions and a quicker response, that might not entirely correspond to the CoP but which are however delivering best possible information with the highest possible doses of quality assured by official statistics. For this to happen, a major mind shift of all actors involved seems to be a precondition.

In conclusion, for the above, knowledge and expertise do exist at national, European and at wider international level, also structures for discussion do exist. What is needed now is a commitment from various actors (on official statistics in general, and health statistics and epidemiology in particular) to review the available data, information and experience as well as the actions taken in the current situation, and to fill the gaps in type and quality of data.

May 2020
Dr. Marleen De Smedt
Former Head Eurostat, Unit ‘Health statistics’ Former Adviser to DG Eurostat
------------------------------ 10

Hasnae Fdhil

Comment on:
04/06/2020 12h36

Impacts of COVID-19 through household surveys

As stressed by many comments, the role of the NSOs in enlightening decision makers and public debates during and after this unprecedented pandemic is crucial. The NSOs are involved in a global, regional and national dynamic to adapt the existing surveys & data sources and to build new ones to assume their role.

I would like, through this contribution, to share with you the results of a household survey conducted by the High Commission for Planning (Morocco NSO) to measure the impact on the social economic and psychological situation of the Moroccan households. The survey was conducted by phone in early April on 2350 households and covered different items. I’m giving bellow an idea on the main results but more details can be found through the link:


1. The level of effectiveness of the lockdown reveals, among others, that 80% of the population fully respected it whereas 20% did it partially. Domestic supply is the main reason for leaving home (94% of households) while work is the reason for 30% of them.

2. Supply of consumer and hygiene products: While for almost all households, the supply of basic food products and butane gas remained normal during the lockdown, the supply of hygiene products was contrasted, as for instance, 33% of households have sufficient protective masks, 41% have them in insufficient quantities and 27% don't have them at all. The survey showed also discrepancies between poor and better-off households.

Measures were taken by the government to reinforce the local production of the masks.
3. As for the impact on the household’s income, 34% of households claim to have no source of income due to the cessation of their activities at the time of lockdown. One in five households has received state aid to compensate for job loss.

4. From the education perspective, seven out of ten households are moderately or not at all satisfied with the channels used by their schooled children. The lack of interactivity is the main reason for the dissatisfaction, as stressed by the households.

5. The lockdown impacts also on the population's access to health services, thus among 30% of the population concerned by chronic diseases, 48% have given up on health services mostly because they are afraid of being infected with the virus.

6. Finally, it should be noted that the lockdown have impact on family relationships too: 18% of households felt a deterioration in family relationships.

It’s clear that producing such data timely helps in addressing public policies as for instance the survey shows that Radio and television are the main sources of information on the pandemic (87%) while social media is the source for only 6 % of the households. This data is relevant to orient the Covid-19 protective campaign through these channels. From another perspective, knowing that 36% of the children eligible for vaccination had to give up immunization services, and 30% of women eligible for antenatal and postnatal consultations had to give up these services during the lockdown, would be useful for health department to take concrete measures in this regard.

A second round of this household survey is planned in June to inform on the evolution of the impact on all individuals to track discrepancies inside the households (gender, aging, employment, etc.) and to introduce new questions on the exit challenges.


Comment on:
01/06/2020 15h39

Rapid action and substantial investment are needed

The effects of the COVID-19 crisis have hit severely some national statistics offices (NSOs). A metamorphosis --understood as a rush of irreversible, unstoppable, and involuntary changes-- is taking place in the traditional statistical methods affected today by the suspension of face-to-face data collection. Although NSOs want to measure the impact and prevent the generation of statistics from falling, reality prevails.

An example is the 2020 Population and Housing Census (the 2020 Census) carried out in Mexico by the INEGI. It was possible to complete the mass face-to-face enumeration stage, and remote monitoring of about a million invitations continues. Inevitably, some processes appear affected.

Other examples are the telephone interview survey methods that INEGI is already testing for the occupation and employment (Encuesta Nacional de Ocupación y Empleo, ENOE, in Spanish) and consumer confidence (Encuesta Nacional sobre Confianza del Consumidor, ENCO) surveys. Looking for alternatives to get through the social distancing that prevents traditional data collection in the survey related to household income and expenses (Encuesta Nacional de Ingresos y Gastos de los Hogares, ENIGH) is underway.

These difficulties made it impossible to continue according to the original plan. Although INEGI is working on tactics to focus available resources on alternative solutions, a pending issue is to confirm if our creativity and disposition are effective.

Challenges include changes in sample sizes and, in some cases, scheduled dates. This directly affects the time series we built over the years, complicating the comparability of results. The impact is happening now, as we write these lines, and our budget did not include coping with such events.

The current situation could justify the request for additional resources in the federal budget that the Mexican government allocates to INEGI year after year. It could even justify the request for resources from international organizations. However, the funds would be available until 2021 and they will not be able to correct the impacts that we are living now, in 2020. In this regard, INEGI has more questions in common than answers with other NSOs.

Notwithstanding, NSOs must continue promoting the importance of national statistics production with governmental authorities and international organizations. Only data that carefully describes reality is helping decision-makers solve challenges posed by the COVID-19 pandemic. As well, quality data will be key for the tests that we will face in the construction of a new society once the early stage of this crisis ends.


Comment on:
01/06/2020 15h38

Statistics will play an active role in shaping the world of tomorrow

It is time to take advantage of the crisis to make national governments aware of the importance of having sufficient and timely statistical information so that national statistical offices be strengthened.

It will be very important to continue working proactively to place statistics at the center of the public agenda, so that most of the strategic decisions that governments make in the current crisis are based on statistical knowledge, and that in the future, public policies designed and implemented will have them as a reference.

An important lesson for national statistical offices is the need to widen the sources of information and make them sufficiently reliable. An important lesson for governments is to work hand-in-hand with their statistical offices to generate strategic information in shorter time frames.

In Mexico, the generation of statistics is regulated by the National System of Statistical and Geographic Information and its Law; this establishes cooperation mechanisms between the public, academic, private and social sectors for the generation of statistics; however, it is necessary to create new mechanisms or to strengthen the existing ones; this will be fundamental to improve the opportunity to obtain information from sources other than the usual ones, such as that which could be obtained from telephone directories or e-mails from registers or users, as well as the more efficient use of administrative records.

The challenges imposed by the COVID-19 have forced the rapid search for ways to produce statistics suitably, breaking important methodological paradigms; however, these new forms are not being as efficient as required in terms of representativeness and quality of the information obtained, not only because of the problems of alternative sampling frameworks but also because of the high levels of non-response and distrust of the population in the face of new data request methods. All of this will have an important impact on future activities, even after the confinement stage is over, in which regaining society's trust will be one of the greatest challenges.

The NSO´s must be attentive to the information needs that arise after this crisis. The economic and social impact that derives from it could impose new necessary indicators for the decision-makers, it is highly probable that the perspectives with we approach the reality will change because it will be different from the one that was had before the pandemic. It is also very likely that the priorities and goals of governments will be modified, an example of which is the importance that the Mexican government wants to give to the measurement of happiness beyond the traditional economic indicators.

Placing statistics at the center of the public agenda requires a quick response to the needs of both national and international decision-makers.


Comment on:
01/06/2020 15h37

Official statistics’ has to adapt quickly to the changing circumstances

Many factors have influenced the development of statistics, including scientific and technological progress. The health emergency caused by COVID 19 substantially modified our environment, so statistics must adapt to new circumstances.

The impact of the COVID-19 pandemic on the formulation of official statistics

The impact is already being seen in official statistics, as it had to suspend or postpone some census activities and several surveys. This is the reason why the actual statistical time series will not be maintained and their quality will likely be affected.

To replace them, changes will have to be made in the conceptual and methodological frameworks of the current statistical programs.

The role of NSOs in the era of the data revolution

Just as large private companies use data analysis to make better decisions while minimizing risks; Governments at all levels should do the same.

Every crisis carries with it areas of opportunity. Perhaps this is the right time for NSOs and governments to work together. On the one hand, to strengthen statistical products, including administrative records and, alternative data sources. On the other hand, the statistics generated by the NSOs are a valuable tool for making assertive decisions by governments, which allows calculating and minimizing risks, reducing asymmetries and making projections to improve public policies. The production of statistics by the NSOs must be a priority and strategic activity, to overcome as quickly as possible the problems caused by COVID-19.

The most important challenge is that the different government offices at all levels coordinate institutionally with the NSOs in the production and use of the data for the benefit of the country. For example, applying derived statistics to administrative records could create new statistical products to be used in government decision-making.


Comment on:
01/06/2020 15h36

A new role for official statistics in the public statistical infrastructure

In Mexico, the INEGI is an autonomous and innovative organization. However, the COVID-19 imposes new challenges that even with these characteristics must be considered.

Alternative statistical exercises can be carried out, given the restrictions imposed on traditional methods. However, limitations have been observed, for example, the self-enumeration strategy in the case of the Population and Housing Census 2020 (CPV 2020), for people to self-enumerate through the Internet or by phone, has not had the expected results; in the case of alternative strategies to capture information by phone, they face a certain level of distrust from the population to respond, in addition to not relating this type of information collection with the ONE of Mexico (the INEGI).

As for the alternative of telephone requests and letters for self-registration, people have not wanted to share their personal e-mails, so they have resorted to telephone directories, as well as to search for people's addresses through geo-referencing and in administrative registries such as the national dwelling cadastre, which still needs to be perfected, using other databases, to provide greater precision. However, in view of the low levels of response through these methods and in some cases the small number of samples with which we are working, there is still a need to continue looking for alternative channels of communication to obtain some information from informants, since there is no complete telephone directory or household cadastre.

Regarding the sample framework that is affected during the confinement stage, alternative solutions are being developed to make the housing samples available for some of the INEGI surveys; however, once the health contingency is over, the work of the Master Sampling Framework should be resumed under the same conceptual, statistical and methodological standards of the statistical program.

The new data collection methodologies, and the use of statistical estimation models, such as those for small areas, will be useful. However, the crisis emphasizes government coordination and motivation to have valuable information to overcome the pandemic in all areas, health, economy, education, among others. The new challenges are focused on the mistrust of citizens to respond to surveys that give us the necessary information.


Comment on:
01/06/2020 15h35

Official statistics should be on board to help in crisis management

Statistical information should be shared so that decision makers have input in the context of IDRC-19. However, governments should learn that quality statistical information serves to improve decision-making and overcome the challenges posed by the pandemic, and efforts should be focused so that NSOs have the support required to operate in the new context.

In the case of Mexico, the INEGI is preparing the COVID-19 Digital Map: context indicators containing information on the population groups, with emphasis on the population aged 60 and over, number of people suffering from chronic diseases related to COVID-19 deaths, number of deaths from related chronic diseases, national database of hospitals and clinics, density of hospital beds per 100,000 inhabitants, among others.

Mathematical models in health, economics and employment, among others, can be used to calculate risks and predict future scenarios in these areas. Likewise, new ad hoc surveys can be generated on the situation and strategies to capture information that cannot be obtained with traditional methods (teleworking, telephone interviews, and derived statistical models). However, it is stressed that governments will have to pay attention to priorities, since in times of pandemics statistical information is crucial, and should be a fundamental axis in parallel with health measures.

The continuity of time series, employment rates and other statistics of interest should be sought, including expanding the use of administrative records for the generation of statistics. It would also be necessary to share the experiences of the NSOs with other organizations and government agencies, as well as to learn from international good practices.


Comment on:
01/06/2020 15h33

SDG’s information as the all-encompassing framework

Mexico has made significant progress consolidating statistical information through a regulatory framework that grants autonomy to its national statistics office and regulates the generation of national interest information. The monitoring and measurement of the SDG´s have played a fundamental role in the development of statistics, as well as products for their dissemination.

With the support of the National System of Statistical and Geographic Information Law, progress has been made in the generation of a methodological framework for monitoring SDG´s through a Specialized Technical Committee on SDG´s in charge of their measurement. The Committee is presided by the Government of Mexico and its Technical Secretariat is headed by the INEGI; besides, all the state ministries and government agencies involved in compliance with the SDGs participate in it. In this way, in addition to being involved in the compliance of goals, they participate in the generation of the indicators and know the impact of the measures implemented at first hand.

The current context makes it even more difficult to generate administrative records due to the lack of personnel considered non-essential, but it is necessary for some national governments to reflect on whether or not the generation of statistical information is indeed an essential activity. Up to now, administrative records as a source of statistical information present many shortcomings in developing countries, since the methods for obtaining them are archaic; it is urgent to change this situation in order to be able to make use of them, promoting automation in their collection. This experience of crisis should be useful to national governments to find the way to make efficient, homogeneous and obligatory to make available to National Statistical Offices their records captured with quality and delivered in a timely manner, for which the NSO´s could advise public institutions.

To solve the possible lack of information, INEGI relies on strategies such as the Generation of Statistics for Small Areas, which combines information from household surveys, censuses and administrative records, which helps estimate important indicators in contexts of social confinement, such as the current one. Other techniques and mathematical models have been tested and used to provide the necessary information to Mexican society. The INEGI has been working for some years in the generation of open data and in the follow-up of the SDG´s in collaboration with the government where the key element is the determination of information of national interest. The information is available in the following platforms: http://agenda2030.mx/ODSopc.html?ti=T&goal=0&lang=es#/ind https://www.inegi.org.mx/servicios/datosabiertos.html

Although progress and strengths have been made in monitoring the SDG´s, the current situation poses many challenges, including maintaining the generation of indicators despite the government's halting of face-to-face interviews, achieving representativeness with different methods, and maintaining similarity of traditional indicators with those captured through new methods.


Comment on:
01/06/2020 15h31

Not just the current, but many new statistics are needed

The world, as we knew it before COVID-19, will change in many activities of daily life. Official statistics will be no exception. For this reason, it will be necessary to maintain, as much as possible, the quantity and quality of the statistics currently produced. Once the crisis is over, the creation of new statistics and methodologies will be necessary to quickly obtain the data that allows measuring the impact of COVID-19 on the economy, health, education, etc. Both challenges are not minor.

What will the new statistics reflect/say?

In the case of the new statistics that will be necessary to measure the impact of the pandemic once the health crisis has been overcome, it will be essential to develop an entire methodology that allows them to be obtained, analyzed and published as quickly as possible and that also reflects the reality of the country. From the design of the surveys to the publication of the results, difficulties arise that must be overcome. What to measure and how, how to select the samples, how to interview people (face to face, by phone, etc.), how to get them faster and be reliable, how to interpret them, among others. Clearly, methodological changes imply changes in interpretation. To solve this problem, new technologies (hardware, software, methodologies, etc.), the experience of ONEs, international cooperation via knowledge exchange, and the support of institutions such as the World Bank or the Inter-American Development Bank will have to be used. An example of the latter is the regional high-frequency survey to measure the socio-economic impacts of the Covid-19 pandemic sponsored by the World Bank.

Another alternative commonly used by developed countries is to make use of administrative records. In the case of Mexico, it is not so simple, since the methods to obtain them are archaic, as they depend on person-to-person contact, in addition to the problem that this confinement represents to obtaining them since only essential personnel are working to avoid human interaction. Hence the need to modernize the system for collecting them by the Government Offices, with the advice of ONE (INEGI) to have automated records similar to those of the financial sector or large companies.

Last but not least, the use of derived statistics to obtain new products from the statistics that are already available would be another alternative way of generating statistics. In the case of INEGI, obtaining statistics at the municipal level based on the state ones, using estimation techniques for small areas from mixed linear models, which solve the problem of having a single point per municipality, when needed at least two to build the line from which municipal information is obtained.

Everywhere you hear about statistics without being completely clear if the transmitters and receivers understand them, so it would be advisable to educate people about and improve the way of communicating them, in such a way that people really understand them. After all, we are in the midst of the data revolution, and it would be desirable for ordinary people and the government to make better use of all available information to make better decisions, just as large private sector companies do.


Comment on:
01/06/2020 15h30

Now epidemiology is most important, in the next phase other statistics will become relevant again! The enormous importance of statistics in the preparation of difficult political decisions has seldom been as clear as it is now in the current phase of the pandemic crisis.

The world has not yet overcome the so-called “first phase of crisis management” of the COVID-19 and some national statistical offices (NSOs) are already experiencing the rigor of its effects. In developing economies, no office has been known to say that it is ready for the new normalcy once said phase is over.

Redesign of face-to-face data collection methods is inevitable if the ability to generate statistical information is to be maintained. However, reality is imposing hard lessons that will lead NSOs towards a methodological metamorphosis.

For example, in Mexico, surveys related to occupation and employment (Encuesta Nacional de Ocupación y Empleo, ENOE, in Spanish) and consumer confidence (Encuesta Nacional sobre Confianza del Consumidor, ENCO) must now continue with different survey methods applying telephone interviews. Other surveys, such as the one related to household income and expenses (Encuesta Nacional de Ingresos y Gastos de los Hogares, ENIGH), are also looking for alternatives to overcome the social distance that prevents the traditional data collection.

The new methodological designs should incorporate changes in the size of the samples; the initial trend seems for them to be reduced. However, the greatest challenge will be the success of contact with the target population, the basis of data collection. For this, it will be necessary to integrate social and technological factors along with features resembling ad campaigns and customer care services to attract and keep sample participants.

Not only will (i) contact directories with updated personal data be necessary. It will also be vital to incorporate ways of generating credibility and trust based on elements such as (ii) the reputation or social acceptance of the NSO and (iii) mechanisms that allow the public to detect and rule out data collection apocryphal campaigns.

The risk of those new methodologies and sample sizes is that the scope of NSOs will become like that of market research firms working today. Reality will need us to be creative so that, after the changes, official statistical information continues to be useful and unique for the design of public policies that advance all sectors in the new normalcy that will prevail.


Comment on:
14/05/2020 18h50

Official statistics should be on board to help in crisis management

In the near future, politicians will be confronted with extremely complex decision problems. The technical consulting teams for crisis management will therefore be multidisciplinary. It would be of major importance for statisticians to be on board here to contribute their various potential strengths, such as wealth of methods, experience in various applications, comprehensive information infrastructure, long time series, solidity and trust.

In reaction to this statement, I am happy to announce that yesterday (May 13th) the CCSA (The Committee for the Coordination of Statistical Activities) - 36 international organisations came together and published a special report "How COVID-19 is changing the world: A statistical perspective" to try and address this exact issue. You can find this report at: https://unstats.un.org/unsd/ccsa/documents/covid19-report-ccsa.pdf

For those unfamiliar with the CCSA, it is comprised of international and supranational organizations, whose mandate includes the provision of international official statistics in the context of the Principles Governing International Statistical Activities and which have a permanent embedded statistical service in their organization and regular contacts with countries. The mandate of the CCSA is to ensure the efficient functioning of the international statistical system; develop common standards, platforms and
methodologies; provide inter-institutional support; outreach; and advocacy for high quality official statistics. More information can be found on the CCSA webpage: https://unstats.un.org/unsd/ccsa/


Dennis Trewin

Comment on:
02/05/2020 14h23

Last weeks I facilitated a workshop between statisticians and some of the epidemiologists and bio-mathematicians working closely with Government on the COVID-19 crisis. The objective was to work out how statisticians could best assist their endeavours. Not all are directly related to official statistics but I thought they may be of interest.

IN THE SHORT TERM (ie over the next month)
1. The tests to date have been on a ‘self-selected’ basis. They will not be representative of the population at large and so care needs to be taken on what inferences can be made about the population at large. For example, NSW and Victoria are encouraging testing to anyone who wants it. This will have a bias towards those people who have the time and capacity to drive to a site. Young adults and children are less likely to be tested. The better educated are more likely to be tested. Statisticians could provide advice on how get the best possible population inferences from the active virological (self-selected) testing regime (by geo-demographic reweighting?) and the likely limitations (biases). They could explain what the estimates mean (e.g. possible upper bound rather than actual ratio of infected persons). The extent of advice depends on the amount of geo-demographic information that is associated with individual test results.
2. At first glance, the self-selected sample might seem to produce an upward bias because people who are symptomatic are more likely to be tested. However, it is also likely that the self-selected sample will under-represent children and young adults especially males. These may be more likely to be asymptomatic or mildly symptomatic although not tested and, if so, this would provide a bias in the opposite direction. Statisticians could provide advice on how to best supplement the active testing sample with populations that will be under-represented.
3. Statisticians could Review procedures for estimating R0 (the basic reproduction number). This is a crucial statistic for future planning but the statistical science behind it might be able to be improved (Methodology available on the Doherty web site).
IN THE MEDIUM TERM (Until the virus is under control but should start work in the short term)
4. As soon as feasible, a national survey be conducted jointly by the ABS and health authorities to estimate the number of infections after restrictions start being released. This information is very important for monitoring the potential for a second wave that is difficult to contain. It is may become more important in the winter months when there will be more overlap between COVID-19 symptoms and other winter ailments. Also, the number of asymptotic infections is unknown. International studies show the percentage is much higher than what health experts believe is the case in Australia. Statisticians might be able to assist the ABS with the analysis of the data. The survey should be repeated until there is confidence the virus is under control.
5. Assist with the design, calibration and analysis of sewage testing. It could be combined with the national survey in the first instance to provide a ‘calibration experiment’. This would allow calibration with what we see in the population reducing the need for large scale surveys in the future.

M. De Smedt

Comment on:
29/04/2020 18h52

Thank you, Walter, for launching this very timely discussion ! The points you raised in your text are indeed the key issues which need attention not only from the statistical community, but also from politicians, managers and the public at large.

I would like to contribute with the following reflections:

1. Health matters

Having spent a significant part of my career in the field of health statistics at EU level, I had to convince many colleagues and superiors of the importance of the domain (see comment of M.Wolfson). The Covid19 crisis now clearly demonstrates that those who say that health is only a matter of a patient and his/her doctor and that statistics on ‘health and safety’ are not important, have to think twice.

2. All domains and sectors are affected by this crisis

This Covid19 crisis has also dramatic consequences on many other aspects of our personal lives and on so many sectors of our societies at local, national or global level. This type of worldwide crisis needs a comprehensive management and monitoring infrastructure and an active engagement of many actors in all domains and sectors and at all geographical levels.

3. Managing the health crisis, scientific expertise at the front level

As for the immediate management of the impact of the SARS – cov 2 virus on public health, all actors as well as the larger public are eager to see the daily facts and figures. But, beyond figures, more information is needed and available (as mentioned by J.Dunne), such as quick results of relevant studies and research on the virus, results of modelling and expert opinions on different confinement measures and exit strategies and on possible medications and planning for a vaccin.

Efficient cooperation between different providers of information and data (statistical institutes, universities, ministries of health and national scientific institutes on public health) is essential. The statements ‘public health is the first priority’ and ‘we rely on scientific expertise’, are now heard in many of the governments’ press releases, but also in many daily conversations.

4. Official statistics; about data and statistical tools

Because of the urgency, countries initially relied on rapidly available (health) data, which do not necessarily fulfill the quality requirements of official statistics. As mentioned by some of you in your comments to W.Radermacher's text, several national (and regional) statistical institutes have stepped up efforts by providing more rapidly official health statistics in order to correct and complete the early figures. They also offered their statistical expertise on methodologies, reporting mechanisms and quality assurance for developing new instruments. As mentioned by P.Diaz, providing such services of official statistics, has resulted in the ad-hoc surveys on Covid19 (launched or about to be launched in BE, ES and FIN) destined to fill in the data gaps, mainly on the impact of the virus and of confinement measures on people’s lives.

5. Monitoring and managing all aspects of the crisis

Good information and data on so many other sectors and domains is needed, as well as guidance on the objectives to be achieved at all geographical levels. As W. Radermacher states, the Agenda 2030 provides the worldwide framework on the way forward, on the (17) SDGs and targets to be achieved and on the monitoring required. important to know is that not only national and local governments, but also many companies and NGOs are using the 17 SDGs as a framework to set their own goals.

This is good news: the framework is there and could help us in monitoring and managing us in the way out of this crisis.

6. What could/should be done at EU level ?

The competences of the EU in the field of public health are now solely limited to exchanging information, promoting public health and helping countries tackling common challenges such as in the case of pandemics. But organizing and delivering health care is the responsibility of the national governments which explains the different approaches and measures taken by countries in this crisis situation. Only for health statistics, a statistical legal ‘package’ is available.

At a later stage, health statistics - building common definitions and methodologies at EU level - will show us the outcome of these different approaches and could maybe lead countries to give the EU more competencies in the field of public health.

A more detailed article will follow.

Marleen De Smedt, former Head of Unit on health statistics in Eurostat and former adviser (on SDGs) to the DG of Eurostat


Comment on:
28/04/2020 12h19

A government roadmap for addressing the climate and post COVID-19 economic crises (https://climateactiontracker.org/publications/addressing-the-climate-an… )

"The COVID-19 pandemic presents the world with an unprecedented policy challenge: not only will it have a severe impact on the global economy likely to exceed that of both the 2008-09 Global Financial Crisis and the Great Depression, it will take place against the backdrop of the ongoing climate crisis.
In acknowledging the magnitude of this unprecedented challenge, the priority for governments must first be the immediate emergency response focussing on saving lives, supporting health infrastructure, food availability, and the many other urgent social and economic support measures such as short-term job allowances, direct cash handouts to citizens, or targeted liquidity support to SMEs. But solving the COVID-19 crisis cannot come at the expense of solving the longer-term issue facing humanity: the climate crisis.
The question of how the economic recovery is designed remains crucial in shaping the long-term pathways for emissions and determining whether the Paris Agreement’s 1.5˚C temperature limit can be achieved. Our analysis points to strong economic and climate change advantages if governments were to adopt green stimulus packages in response to the COVID-19 pandemic.
Conversely, if governments don’t roll out low carbon development strategies and policies - or roll back existing climate policies - in response to the coming economic crisis, emissions could rebound and even overshoot previously projected levels by 2030, despite lower economic growth in the period to 2030. This report shows that the future is for governments to choose. COVID-19 recovery presents both opportunities and threats to enhancing our resilience to climate change."


Comment on:
25/04/2020 22h59

I find the discussion and opening statements very interesting
I would like to make a sort comment in the context of the following two statements

- Official Statistics should be on board to help in crisis management

- A new role for official statistics in the public statistical infrastructure

To start my comment I would summarise the role of a National Statistical Office (NSO) briefly as
"ensuring the right of a person to live in an informed society ", balancing this role with the right to privacy while undertaking this role in an open and transparent manner. Informed decision making is necessary for smarter and better outcomes.

In fulfilling this role NSOs have developed and implemented relevant infrastructure, methods and standards with respect to their work. One primary competence is safeguarding data.

Much of the subject matter knowledge to inform the Government response to C19 typically resides in academia type settings or other specialised teams, not necessarily in NSOs. Much of the data needed to inform the Government response lies in Government or privately held databases . This gap can cause difficulties for many countries.

In times of crisis, is there a new role emerging for NSOs (and ISOs) to react quickly and safely in partnership with stakeholders to ensure relevant decision makers have fast and reliable information for smarter decisions and better outcomes. Is this role already here - I think it is .... NSOs need to develop the capability to move fast, if they haven't already.

John Dunne


Comment on:
23/04/2020 15h54

About the way forward I second Walter when saying "Rapid action and substantial investment are needed".

Advocating, raising awareness about the situation of Official statistics are permanent activities I would say. However I guess many in the health and other sectors will probably do the same and governments will soon be overwhelmed with urgent claims.

Sure "international networks and cooperation" should be promoted and reinforced. However I believe much more is needed if the shared aim it to embark all countries towards achieving SDGs on or before 2030. Rallying around the Cape Town Global Action Plan for Sustainable Development Data, internationally agreed upon in 2017, is needed to globally succeed, each country strategyzing its own path to measure the SDGs indicators, although we already know that more than just a few countries will not be able to measure all indicators. Strategyzing support for capability building directly linked to the delivery of the needed data is indispensable.

2030 is just a step, agreeing on a mission and a vision for Official statistics nationally and globally, should force us to consider a longer perspective (2045?).

Gérard Chenais


Comment on:
22/04/2020 19h27

My experience is mainly with official statistics in small and not big developing countries (Pacific and Africa) and I was faced all the time with this that Walter says "there is no such thing as the quasi fateful definition of what official statistics means". To me, it is not really a problem when dealt with between national official statisticians in advanced countries, much more when those providing advices and support, technical assistance and/or finances, to improve the national statistical system of developing countries have strong but differing views on what it is about, and on what they expect for themselves. The design of a national strategy for the development of statistics is then a rather difficult exercise; country ownership is not spontaneously adhered to. Already in the previous comments one can identify Official statistics as a profession (those working on official statistics mainly in the public sector) or as some kind of useful data. I would prefer to consider it as a public service function delivering statistical data as public goods; the question is who to, the answers lead to quality requirements that serve best decision-making.

Here is what ISO expressed about evidence-based decision making:
• Make all data needed available to the relevant people.
• Ensure that data and information are sufficiently accurate, reliable and secure.
• Analyse and evaluate data and information using suitable methods.
• Ensure people are competent to analyse and evaluate data as needed.
• Make decisions and take actions based on evidence, balanced with experience and intuition.
This sounds obvious to most of us, the last point is important but out of our technical sphere!

Coming back to the news issues that the Covid-19 imposes to statisticians, I am very much confident that the Big Statisticians will provide the best possible responses at short and longer terms ; Government statisticians in charge of the systems of developing countries would need the best possible advices from Big Statisticians to design their own strategy to cope with these issues, in the mean time decisions makers lacking the useful evidence will have to decide and take actions based on their experience and intuition, as they are used to.

Gérard Chenais


Comment on:
22/04/2020 17h40

In the age of SDGs national statistical systems face a much bigger pressure than before as they need to deliver many new indicators, which require new data of high granularity to be able to adhere to the principle “no one left behind”. All countries face constraints (capacity and financial) to measure achievements of SDGs as agreed at the highest political levels.
COVID-19 puts new pressures on the national statistical systems as the response requires high frequency data from different sources--NSO, health monitoring systems, administrative and private sector data--to make sure databases are coherent.
Most importantly, the data must be presented to the policy makers, health care professionals and to the public at large in open formats to boost trust in data.
What we see in the USA, every new figure presented by the official statistics, are questioned by the media, so are the models which use the data. That creates a chaos in the minds of ordinary people who need to trust and believe in something, particularly in difficult times. Open formats with metadata explained in a simple language would help a lot.


Comment on:
22/04/2020 16h38

A pandemic is not the only event or instrument which is drawing official statistics out of its national borders. As Walter Rademacher observes, the SDG data framework de facto gives it a global context. Likewise the crisis in biospherical evolution demands cross border monitoring if effective action at whichever level is to result. The methods training of statisticians for work within NSOs is to a large extent neutral as to territory or jurisdiction of the data production. What is lacking at global scale, as commentators have made eloquently clear, is authority to design and construct the global systems of statistical production to meet these information needs - not an absence of funds but of vision. Global official statistics is more than an idea floated in a journal, yet has a shaky foundation. We have the policy setting - via the SDGs, or via some other concensual frame.

We may now have the capacity to cooperate in carrying forward the construction - granulated or on a bold scale: by analogy the impetus for methods advances from the 1950s that shaped the current industry;. approaching 2030 the moral landscape demands attention by contrast to the value of life in a deteriorating and unsafe environment, not one of rising (but uneven) prosperity. This Dr Rademacher has admirably challenged the official statistics community to engage with