Country practices
The examples below illustrate how NSOs deal with the challenges of continuing their operations during the COVID-19 crisis.
The main sources of information are: (1) a special survey carried out by the UNECE Task Force on Measuring Hazardous Events and Disasters (and follow-up communication with NSOs), (2) the UNSD COVID-19 response website, (3) NSOs' websites and (4) information provided directly by NSO.
For updates and adding of information please leave a comment at the bottom of the page or contact Malgorzata Cwiek.
Country | Category | Title and/or link | Description | Source | Date added |
---|---|---|---|---|---|
Australia | Consumers and data suppliers management | Australian Statistician's Briefing to the Australian Business Economists | ABS has created three webpages dedicated to COVID-19 response:
On 30 April, the Australian Statistician gave a briefing to the Australian Business Economists on:
| 3 | 11 May 2020 |
Australia | Statistical production | ABS COVID-19 statistical products Australian Statistician's Briefing to the Australian Business Economists | The ABS is committed to delivering timely and accurate information about the impacts of the coronavirus (COVID-19) on Australian society and the economy. The ABS is producing a range of existing and new statistical products, providing relevant insights on households, employment and industry to inform government, business and community responses to the pandemic. A dedicated task force has been established to identify and secure new data sources to supplement the production of existing ABS products, address emerging policy questions and data needs in response to COVID-19, and deliver novel and innovative products in the future. New statistical releases will provide information on:
| 3 | 11 May 2020 |
Australia | Statistical production | Protecting the health of staff and respondents, while ensuring the business continuity | ABS is protecting the health of its staff and respondents, while ensuring the ongoing operating capacity. By mid-March, given the risks associated with travel and gatherings in confined spaces, ABS reduced the number of agencies and individuals attending lock-ups for market sensitive statistics. ABS also suspended surveys that required close face-to-face contact, including those within Aboriginal and Torres Strait Islander communities, to protect vulnerable populations. Some of the regular activities were paused to free up resources to ensure the key economic series weren’t at risk, or to make way for the collection of COVID-19 related data. Critical teams were identified, and surge capacity and shadow teams put in place to keep production processes going in the event that infection rates of ABS staff rose to critical levels. | 3 | 11 May 2020 |
France | Statistical production | Impact of COVID-19 on statistical production (press releases of 16 March and 25 March) and a dedicated COVID-19 web page | INSEE has been complying with the confinement and has been able to ensure continuity of its mission through teleworking. This required redefining priorities and adapting certain operation to the new conditions. INSEE informs about the impact of the COVID-19 on its work through press releases and a dedicated web page. Registers: INSEE assured the public about that it will continue to maintain the population register and the business register which are under its responsibility. Surveys: INSEE carries household surveys exclusively by phone or online if possible (employment survey and living conditions survey). ICT survey will not be carried out in 2020. Replies already provided by companies online will be preserved but there will be no reminders and no legal consequences of non-response. Certain large companies whose responses are indispensable can be contacted. Statistical production: INSEE aims to ensure the continuity of the production of national accounts. The provisional annual accounts for 2019 should be published as planned. The publication of trimestral accounts for first and second trimester 2020 may not follow the usual calendar, in particular the first estimate, which is scheduled 30 days after the end of each trimester. INSEE will continue to publish according to the usual calendar (or almost) all the main economic indicators, including monthly industrial production index, trimestral employment figures and monthly price index. Economic developments and difficulties in data collection may however affect the quality of certain indicators. This will be systematically documented. Every week, INSEE will publish daily number of deaths reported by the communes, based on the civil registration system. INSEE is in contact with mobile operators to measure the distribution of the population on the territory. This work will lead to a publication as soon as possible. Economic impacts: INSEE aims to publish an analysis of the economic situation every two weeks. INSEE will issue updated press releases as necessary if conditions of data collection deteriorate or if INSEE encounters difficulties in ensuring continuity of service. | 4 | 4 May 2020 |
Netherlands | Statistical production | A dedicated website Coronavirus crisis: CBS figures | Statistics Netherlands created a dedicated website with CBS figures on COVID-19 impacts on economy, society, public health, public mobility and supply chains, and related news articles. | 4 | 4 May 2020 |
Netherlands | Statistical production | Measuring article stockpiling with scanner data | On 24 March, Statistics Netherlands published an analysis on article stockpiling based on weekly transaction data. | 4 | 4 May 2020 |
Netherlands | Consumers and data suppliers management | Message to users of CBS data and those providing data to CBS | On 26 March, Statistics Netherlands published a message to users of CBS data and those providing data to CBS that:
| 4 | 4 May 2020 |
Netherlands | Strategy and leadership | Interview with CBS Acting Director General Bert Kroese on reliable data in times of corona and beyond | On 22 April, Statistics Netherlands published an interview with CBS Acting Director General Bert Kroese, who further explained the impacts of the outbreak on the CBS work, importance of reliable, timely data and the need for continued collaboration between the Dutch national and municipal authorities, the academic world and the business community during the crisis and beyond. | 4 | 4 May 2020 |
Italy | Statistical production | Actions to ensure continuity of statistical production, a special internal task force | Due to measures adopted by the government to contain the virus, Istat has implemented a series of actions to ensure the continuity and quality of statistical production even in this emergency situation. The institute has reorganized data collection by sustainable acquisition techniques, innovative methodologies and use of data sources; it also provided most appropriate solutions to support statistical production processes, in full protection of workers’ health. Official statistics are fundamental for measuring the evolution of economy and society; their production and dissemination at the service of institutions, policy-makers, families and businesses, therefore, cannot be stopped, but need to be rethought to be ready to provide the country with all necessary answers, and above all to support and monitor the future country’s recovery. Istat has long invested in dematerialisation and can therefore ensure full operation of its databases and accessible online services. A special internal task force is dedicated to organize and manage the impact of pandemic on the statistical activities. It caused the stop of some survey and the postponing of others. Istat is working using “agile works” and all the workers are connected and are working as usual. Web-meeting tools are very often used. A special part of the internal website is well-organized containing all the necessary information. | 1 | 30 Apr 2020 |
Australia | Statistical production | Impacts on the work programme | The ABS is providing a range of additional, up-to-date information to enhance understanding of the social and economic impacts of COVID-19, including new information relating to the Australian Labour Market . In order that the ABS can continue to deliver critical economic and social statistics during this time, it has made some changes to some elements of the ABS Forward Work Program. | 4 | 27 Apr 2020 |
Mexico | Statistical production | No interruption of production and publication of short-term economic indicators and maintaining information quality standards | The main challenge is not to interrupt the production and publication of short-term economic indicators and maintain information quality standards. In case of the production of short-term economic statistics, we anticipate problems in surveying and in the imputation methods of missing data, as well as the temporary or definitive suspension of some businesses in certain industries.
| 1 | 24 Apr 2020 |
Mexico | Risk management | Evaluation of the operating conditions in economic units | According to the Contingency Plan of the National Economic Surveys, INEGI will carry out an evaluation of the operating conditions that prevailed during the health crisis in the economic units, in such a way that the data available is ratified or rectified. This activity will be a determining input for the generation of statistical results for the contingency period, in such a way that the data is generated in accordance with the quality, accuracy, relevance and opportunity established by the Law. It is essential to carry out an evaluation of the actions taken during the contingency, both statistical and non-statistical, in order to prepare a detailed report of the actions, highlighting the impacts, conclusions and alternatives of solution for the areas of opportunity that may be found. This report will be the fundamental input for the design of a contingency plan of greater scope. | 1 | 24 Apr 2020 |
Germany | Statistical production | Biggest challenges at the moment | Monitoring of company responses and the share of estimation: Due to the difficulties experienced by companies during the pandemic, non-responses are expected. As far as possible, this is compensated by utilising estimates, accepting a higher error tolerance and reducing the publication depth. The present situation is particularly difficult for many companies, public authorities and private individuals obliged to report data for the production of official statistics. We are nevertheless asking for the submission of data via the familiar electronic transmission channels. Since, this data will help to pro-vide politicians and administrative authorities with the basic information they need to decide on measures for coping with the crisis. Contributions by other public authorities: Important contributions by other public authorities might be missing, because they are also affected by the disease. | 1 | 20 Apr 2020 |
Germany | Human resources management | Biggest challenges at the moment | A reduction of active staff is expected because of the pandemic. The existing staff is "reorganised" according to the statistical priorities. | 1 | 20 Apr 2020 |
Mexico | Human resources management | Providing data to government for crisis management |
| 1 | 20 Apr 2020 |
Turkey | Human resources management | Biggest challenges at the moment |
| 1 | 20 Apr 2020 |
Turkey | Statistical production | Data compilation issues |
| 1 | 20 Apr 2020 |
Turkey | Statistical production | Specific practices | Consumer Price Index: In the context of the COVID-19 crisis; weight structure of CPI and classification which is used in CPI compilation will remain unchanged during the year. However price compilation of the CPI is guided by the following principles:
Short Term Business Statistics: In Short Term Business Statistics, VAT declaration information is used as the data source in the calculation of turnover indices, retail sales indices and industrial production index. Due to the Covid-19, VAT declaration is extended. VAT declaration information of all enterprises has not been received due to the extension of the declaration periods. For the enterprises without data, industrial production survey, e-invoice, e-archive invoice, card payment and foreign trade data were used in the calculation of the indices in addition to the VAT declaration information. There is no problem for the studies whose data source is questionnaire. Household Budget Survey (HBS) Household information is compiled by interview, registration and observation methods in HBS. The households are visited prior to the survey month and the first meetings are held with the Computer Assisted Personal Interview (CAPI) method. Then, during the survey month, each household is visited 4 times for taking expenditure records. After the survey month, the survey is completed by making an income interview with the CAPI method. In this context, if the first interview can be made through a face to face interview it is possible to complete the survey period with CATI method. In March 2020, field application completed by with CATI method. However, field application is cancelled for April 2020 since the first interviews could not be made through face to face interview. Household Labour Force Survey: Before Covid-19, all surveys of the first wave and the majority of the 2nd and subsequent waves were being conducted with CAPI method and a small part of the 2nd and subsequent waves were being conducted with CATI method. Together with the measures taken in the context of the epidemic, all face to face interviews have been stopped. The capacity of the NSO CATI Center has been increased and the CAPI interviewers have started to implement the survey by telephone. Statistics on Income and Living Conditions Survey (SILC) SILC Survey's field application period is 4 months starting at March and finishing at July. This year due to Covid-19 household visits were stopped at the beginning of the field application period. After a while the survey method was changed from CAPI to CATI. For the households of wave 2, 3 and 4 the survey has been trying to be done by phone although SILC is a very long and intense survey. For the first wave households' surveys cannot be conducted and this will be evaluated by Eurostat in the following days. Death Statistics Monthly transferring death data from the data source (the Central Civil Registration System (MERNIS) database of the Ministry of Interior, General Directorate of Civil Registration and Citizenship Affairs) does not allow daily/weekly death data production at that moment. | 1 | 20 Apr 2020 |
Turkey | Human resources management | Main lessons learned | The great importance of technological infrastructure in telecommuting has been understood as it enabled continuation of working process. It has been revealed that the works can be carried out easily by telecommuting and alternately working applications in cases where the personnel health should not be risked, but for this, the existing personnel structure and the technological equipment should be prepared. Flexible working conditions may be applied in the future by evaluating the efficiency of the works. | 1 | 20 Apr 2020 |
Turkey | Statistical production | Main lessons learned | It has been seen how important the coordination with the relevant institutions is for the statistics produced by using administrative records and how important it is to carry out this process with all the stakeholders. Survey Portal (self-completion of computer-assisted questionnaires via Web survey) and EDI (electronic data interchange) Technologies the so called “e-VT” (standard business reporting programs for reporting business data and electronic interchange) is used in the NSO. The survey portal provides respondent units to fill in electronic questionnaires by themselves in a comfortable and a safe manner with the passwords assigned to them. By this system, no disruption in the data collection from enterprises is experienced during the Covid-19 period which highlights the importance of usage of the technological data collection methods. In terms of household surveys, we have used CAPI and CATI methods. CAPI was the main method for household surveys but after a six months pilot study we activated our CATI Center with 30 operators to collect data for some specific surveys. In the coming period, NSO is planning to redesign data collection methods for all its national social surveys in order to employ CATI mode. NSO has established a platform for this task. | 1 | 20 Apr 2020 |
Latvia | Statistical production | Brief flash estimates and new products | Central Statistical Bureau of Latvia has invested its efforts in providing brief flash estimates and new, detailed products in the context of COVID-19:
| 4 | 17 July 2020 |
Latvia | Statistical production | Data dissemination and dedicated section on the web | Data dissemination was adapted timely, so to stay relevant and serve decision makers and public. Extra info graphics were prepared with COVID-19 related statistics, like, map with territories with the biggest share of senior population, number of population with chronic diseases etc. As of 9 April special section on the web page introduced with COVID-19 related socio-economic indicators (number of employees in most affected sectors etc.) All recent releases can be found there. | 4 | 17 July 2020 |
Latvia | Statistical production | New practices for existing surveys |
| 4 | 17 July 2020 |
Latvia | Statistical production | Extra activities |
| 4 | 17 July 2020 |
Latvia | Consumers and data suppliers management | New practices and data sources | Considering that since March 2020 all CAPI interviews were suspended, extra efforts were invested to promote and organize wider use of CATI and CAWI for social surveys. The need for timely and qualitative statistical information was recognized by Cabinet of Ministers and National Parliament, so on 30 April special amendments were made and approved in the Order “On Declaring the State of Emergency”, establishing the obligation for the State Revenue Service, Office of the Citizenship and Migration Affairs and the Road Traffic Safety Directorate to provide upon request from the Central Statistical Bureau the information on contact telephone numbers and e-mail addresses of natural persons included in the sample of the Labor Force Survey, EU SILC and ICT survey. Cooperation with involved institutions has turned out very successful. The work of interviewers’ team is being reorganized accordingly. Besides that, the Central Statistical Bureau of Latvia has been able to agree with the State Revenue Service on faster information exchange on the volumes of performed transactions from VAT declarations, in addition to the existing agreement (letter to the SRS, referring to the state of emergency in the country). | 4 | 17 July 2020 |
Latvia | Consumers and data suppliers management | Communication with user groups | The Central Statistical Bureau of Latvia has put a lot of efforts in communication with user groups during the COVID-19 outbreak. As soon as state of emergency was announced in mid-March, information on suspended home interviews on social media was published and press release prepared. Soon after separate press release for business statistics respondents and individual persons stressing importance of continuous data providing so to assess the impact of COVID-19 was prepared. In the beginning of April special press release with our greatest gratitude to respondents who have continued to provide data for statistical purposes was issued and special call (info graphics) for enterprises and individual persons to participate in surveys followed shortly after. https://twitter.com/CSP_Latvija/status/1258327113754652679, https://twitter.com/CSP_Latvija/status/1258327313852313604/photo/1 (Latvian only) | 4 | 17 July 2020 |
Do you have questions or want to share your experience ? Let us know in the comments below
You need to log in to be able to add comments. If you are not a registered user, please click here to register.
4 Comments
Octavio Rios
I work for the NSO in Mexico (INEGI). We coordinate the Survey on international travellers, in order to obtain the number and expenditure made by tourists from abroad. Since the end of March, face-to-face surveys were suspended, so we don't have the necessary input to estimate data for April. Should you have any suggestions to do so, I will much appreciate your kind guidance on this matter.
Malgorzata Cwiek
Dear Octavio Rios, thank you for your message! As we do not work in the area of tourism statistics, we have shared this question with our network to find some guidance in this matter.
If anyone reading these comments would like to share some experience in this area, please do not hesitate to do so!
Sebastian Wojcik
Dear Octavio Rios . Statistics Poland is the same situation. We used to conduct a survey in the vicinity of the border as well on the airports. At this moment we are working on some experimental methodology assuming that there is no data obtained from interviewees but there are some other data sources which provide update and relevant data to tourism domain. There are two pillars in this approach:
There is a paper titled "National Accounts Estimation Using Indicator Ratios" by Jan R Magnus, Jan W. Van Tongeren and Aart F. de Vos. Authors proposed several methods. We started with quite simple method presented in Lemma A1 because it is easy to code it in R even for beginners. First, we gathered the time series from auxiliary data sources: Border Guard, airports, accommodation establishments statistics from the survey and from the web scraping of accommodation portals, Google Trends. These data sources are available for the quarter of interest. Also we gathered all time series from our survey with correspond to auxiliary time series e.g. trips with respect to purpose, with respect to country (or group of countries), with respect to type of accommodation etc. These data is not available on the quarter of interest.
In this bayesian approach you need to pick priors to all time series. They are produced by ARIMA model or simple trend (depends on the length of time series). Bayesian approach is updating/correcting our prior beliefs for the survey aggregates based on the correlation matrix estimated for all time series as well on the difference between priors and update data for auxiliary time series in the quarter of interest. In that way we get what aggregates we should obtain from the survey database in the given quarter (not all possible but the crucial ones).
Since we do not have a new database from the survey we borrow it from the previous year. Then we calibrate weights using sampling package in R. It implements methods described in Deville, J.-C. and Särndal, C.-E. (1992). Calibration estimators in survey sampling. Journal of the American Statistical Association.
It seems to be a cunning plan but frankly speaking there is a lot of issues. We tried out many phrases in Google Trends to generate anything useful and connected with the purposes of trips. Also, bayesian approach sometimes produced the unacceptable values e.g. negative, thus we tried to start with log transformation of the data at start and log it back at the end. I do not know it is ok any longer... Also I do not know how to compute a precision of estimates.
I am not sure if it can help you. Anyway do not hesitate to mail me if you have any questions.
Octavio Rios
Dear Sebastian Wojcik, thank you very much for your kind answer. I sent you an email providing more elements on this issue. Best regards.