Reliable data in times of corona and beyond

/ Author: Corporate Communicatie
Bert Kroese, Acting Director General of CBS
© Sjoerd van der Hucht Fotografie
Reliable, independent data are a social utility. They have become indispensable, especially in times like these. CBS staff are making every effort to continue their statistics production and inform society to the best of their ability while working from home; in addition, to support central government in tackling the coronavirus crisis by providing new information of relevance. In other words, it is crucial for CBS - especially at this time - to continue collecting data. This is sometimes done by conducting questionnaire surveys and sometimes by tapping into new data sources.

An interview with CBS Acting Director General Bert Kroese, who took the helm as Director General Tjark Tjin-a-Tsoi transferred to Sanquin on 1 April 2020.

Being the Acting Director General in times of corona: what is it like?

Kroese acknowledges these are very challenging times: ‘All of the staff are currently working from home. It is putting quite a strain on people as well as on the organisation. Some of our staff are trying to do their job from home with young children around, who cannot go to school or day care for a while. Others are feeling lonely, missing out on the social interaction at work. Many have concerns about their own health and those near and dear. At CBS, the technical provisions for remote working are excellent, but it is truly a difficult situation for many of our staff.

Yet it is clear that this is a time when the need for reliable and independent data is widespread. We’re working around the clock and under unusual circumstances to meet this need for statistical information. Not only by continuing the existing statistics production, but also by setting up new statistics that are especially relevant now. For example, we now bring weekly updates on company bankruptcies. This calls upon our ability to improvise, our expertise and commitment to new partnerships. We work very closely with organisations such as RIVM, SCP, PBL, CPB and with various private sector stakeholders. In doing so, sometimes we need to come up with creative solutions. For example, surveys that are normally conducted by having interviewers go from door to door, which is simply impossible at the moment. We then turn to alternatives like interviews by telephone and via the internet, so we manage this at least partly.

CBS is still conducting interview surveys. Why not pause these for a while?

Since there is an increased need for data at this particular time, it is even more important for CBS to obtain all the data. The bulk arrive automatically at CBS; nevertheless, we still gather data by conducting written and oral interviews with companies and individual citizens. As of now, one of our core statistics is consumer or producer confidence for instance, it’s very helpful in assessing the economic situation as we speak. This type of data is not available from registers and we really need to find out directly from citizens and entrepreneurs.

Where in society is the crisis hitting hard now, and how hard? We are very well aware of the fact that people now have other things on their minds rather than filling out questionnaires and that businesses are struggling. Yet, we urgently need the data in order to provide the nation with those figures which are particularly needed by policymakers in order to arrive at the best possible decisions.

Could these data not be obtained in a different way?

Of course, we are constantly on the lookout for alternative data sources. We’ve been doing so for a very long time, because we want to lighten the administrative burden on both smaller and larger businesses; but the need to find new data sources seems even more pressing now. However, CBS is acting with caution in this regard. We always consider which data are truly necessary at a particular time (proportionality). Wherever we can, we leave data at the source (data minimisation). We are consistent in our compliance with all the guidelines concerning information security and privacy legislation, these are essential to us. It’s why CBS would never supply data that can be traced back to specific individuals or businesses. And finally, all the statistics we produce are made available to the general public. This is part of our transparency and enables anyone to make use of our statistics.

New data sources: should we be thinking of big data, like mobile phone data?

Yes, for example. We’re looking at how we can support ministries and RIVM in analysing mobile phone data to track public mobility. In addition, we’re looking at ways to accurately present recent economic trends by using data from traffic loops on national roads, parcel deliveries and sensors installed on vessels. We are also studying the possibilities of data on payment transactions such as these are being published by the Dutch Payments Association. With all these data, we want to assist policymakers in mapping (the effects of) the ongoing coronavirus crisis, so they can make the right choices. The prerequisite for CBS is: only if our methods have been screened for compliance with the relevant laws and regulations will we start compiling those data.

What results have you harvested so far?

A fine example is our consumer price survey. In many countries around us, problems have arisen because statistical researchers there actually need to visit retail outlets to measure prices manually. Our survey is based entirely upon big data sources. For instance, web scraping for online shops and aggregated scanner data from physical cash registers. This is how we are able to continuously produce highly accurate inflation figures without any problem. We’re also able to provide an up-to-date picture of consumer spending in shops. For instance, we were able to quickly describe the hoarding at supermarkets recently, at the onset of the coronavirus crisis.

Another good example are our mortality figures. We are now publishing weekly mortality data in close cooperation with RIVM. By comparing these data with the average mortality over similar periods, we can produce an estimate of the number of non-diagnosed deaths due to corona.

Data-driven working is truly in the spotlights now. Do you believe this development is long-term?

Demand for reliable data has been huge for some time. Due to the coronavirus crisis, the urgency to obtain rapid insight and overviews based on reliable, independent data has only become greater. But even around other socially significant issues, we are seeing demand for reliable data grow. Around the world, national governments, the world of science, the business community and national statistical agencies are all innovating with new data sources for national statistics. This development is long-term. Governments are accelerating data-driven working, sharing of knowledge and policy and decision-making. We are making this possible through continued investments in a solid data infrastructure.

Which steps are needed?

There is a great amount of data available within central government; great benefits can be reaped from proper unlocking of this data, so they can be used towards policy development. Data management is no longer merely ‘a side project’. It is a precondition if we wish to manage and boost further developments in the right direction. In an information society in which ‘truth’, reliable data and safe implementation of data have become very valuable, investments in this management and control are needed. Many countries have already taken similar steps in the form of a national data strategy. In more and more countries, a national ‘chief data officer’ and ‘data stewards’ are being appointed. The NSIs often play a fundamental role in this development with their expertise. CBS would therefore like to launch a dialogue with Dutch national and municipal authorities, the academic world and the business community.

Why start discussions on a national data strategy at this particular time?

Especially now in the face of the current crisis, we notice how important it is to have the right types of data to provide orientation. CBS is receiving a lot of requests for quick information from various ministries. Sometimes we need to look hard for the most useful data. By investing further into an open data structure where all public government data can be found and easily accessed, we can respond to such requests more quickly and accurately.

We have also found that new information can be generated by linking and combining data sources. All of this must be under the absolute precondition that we remain compliant with the privacy regulations. This is why CBS puts sharp focus on techniques that enable us to access the data at the source, rather than bring them over. We should then analyse them safely from a distance: privacy by design.

Throughout this crisis, we observe how meaningful data collaboration is between the private and the public sector. By unlocking the data that are available in the private sector in a secure manner, we’re able to create great social value without jeopardising privacy. I already named a few examples. The private sector houses a great deal of knowledge and expertise that can be particularly valuable now. For example, we are now using a web scraping technique in collaboration with a Dutch company to find out how businesses are adjusting in this crisis.

Can you tell us more about the national data strategy?

What we need is a well-functioning system that is operated using solid ground rules. On the one hand, to achieve potential benefits by developing more effective policy measures more quickly. On the other hand, risk prevention: in times like this, matters such as transparency, accuracy and protection of personal data are under enormous pressure. For this, we need to arrive at a general consensus.

A national data strategy for the Netherlands will ensure certain matters are established properly:

  1. Democratic values are safeguarded: ground rules for handling of data are established (think of ensuring transparency, traceability and accountability, privacy, data security and reliability of data and methodologies including algorithms).
  2. Promotion of responsible innovations in the area of data usage and the creation of added social value with the help of data.
  3. Organisation of a technical and administrative (government-wide) infrastructure for data, data sharing and data management. This includes reaching agreements on metadata, standards and harmonised definitions in order to facilitate the exchange of data.
  4. Pooling of know-how and competencies, so working with data happens in the most efficient and secure possible way.

Would such a data strategy only apply to central government?

I foresee an ever closer collaboration between authorities, academic institutions and the private sector, and it is sorely needed. This coronavirus crisis as well shows us that each and everyone can make a contribution. In my view, what the Netherlands needs is a network system of private businesses, academic institutions and public organisations, all collaborating in this fast-developing area of research. Worldwide, data services are growing explosively both in terms of demand and supply, just as the (AI) technology. If the Netherlands wants to keep pace with this growth, such collaboration is actually essential. Private stakeholders have data that can be of great value towards establishing government policies. For example, to find out how sustainable energy targets may be achieved by learning more about household energy consumption. Or to improve local policies on tourism by gaining more insight into Airbnb rentals. On the other hand, there may be businesses that can make a positive contribution by conducting their own analyses and studies. At the moment, we’re receiving a lot of requests from businesses trying to offer their data, but finding it difficult to reach the right government partners. A great deal can be won from this interplay, and we need to shape this further in a wider context as part of a broadly-based societal debate. At CBS we have identified that collaboration with academic and business partners as our own strategic priority.

Data are of vital importance, although ultimately only a means to an end: that is to rebuild a society in which we can live, work and relax together in the same freedom and security we were used to having.