Data sharing

Before migration actors can begin to develop guidelines and procedures for sharing data, they must consider whether the data regulatory environment allows for such collaboration (see Part I, Chapter 3). The actors who are responsible for producing migration data and ensuring data quality may not be clearly defined, national statistical offices (NSO) may not have access to administrative data and private corporations may be unwilling to share data with public sector organizations. Some of these issues may be resolved through the elaboration of a national law on official statistics and the establishment of a national statistical system (NSS) and inter-ministerial working group, all of which are key components of the data regulatory environment. Given that the needs of data users and producers are constantly evolving, developing an efficient and effective internal mechanism for drafting and reviewing new data sharing agreements is important, especially as these processes can take several months.

Protection information management data sharing principles

The United Nations Office for Coordination of Humanitarian Affairs (OCHA) and the Protection Information Management (PIM) initiative, which brings together several UN, NGO and other information management actors, have collaborated in the development of guidelines on the collection, processing and sharing of data informing the development of programmes and actions for the protection of vulnerable persons in humanitarian contexts. The actors involved in the PIM have developed a Framework for Data Sharing in Practice summarizing the minimum shared principles that underlie and characterize the responsible handling, sharing and use of data and information, regardless of their specific purposes, methods or outputs (PIM, 2018). These principles, outlined below, are relevant not only in humanitarian context or for data related to protection and assistance programmes, but also to the production of any kind of personal data.

Table 1: Protection Information Management Principles for Data Sharing in Practice (2018)

People-centred and inclusive

People-centred and inclusive: Data and information activities must be guided by the interests, well-being, and rights of the affected population and their hosts, who must participate and be included in all relevant phases. Activities must be responsive to age, gender and other criteria of diversity as well as the intersectionality of these criteria. 

Do no harm

Data and information activities must include a risk assessment, which looks at negative consequences that may result from data collection, and take steps, if necessary, to mitigate identified risks.

Defined purpose

Given the sensitive and often personal nature of protection information, data and information activities must serve specific information needs and purposes, which are clearly defined and communicated.

Informed consent and confidentiality

Personal information may be collected only after informed consent has been provided by the individual in question, and that individual must be aware of the purpose of the collection.

Data responsibility, protection, and security

Data and information activities must adhere to international law and standards of data protection and data security.

Competency and capacity

Data and information activities are only carried out by information management and protection staff who have been equipped with data and information core competencies and have been trained appropriately.

Impartiality

All steps of the data and information cycle must be undertaken in an objective, impartial, and transparent manner while identifying and minimizing bias.

Coordination and collaboration

All actors implementing data and information activities must adhere to the principles noted above and data and information activities must avoid the duplication of previous activities by building upon existing efforts and mechanisms.

The PIM (2018) refers to three core steps in developing a process for data sharing. The first step is assessing the information landscape from a data sharing perspective, which means:

  1. defining the purpose and potential outcomes of the data sharing exercise; 
  2. articulating the reasons to share;
  3. working together with other stakeholders;
  4. ensuring that the parties collecting and receiving the data demonstrate the required core competencies and respect for the minimum shared principles for data sharing.

The second step is designing the information management system, which means:

  1. assessing the shared data required for decision-making as well as discrepancies in the definitions and concepts employed by different agencies;
  2. conducting a secondary data review to identify benefits and risks surrounding possible data sharing;
  3. reflecting on what constitutes sensitive data within the specific operational context; 
  4. considering the current and future impacts of data sharing on individuals and communities;
  5. ensuring that informed consent was obtained, not only for the specific purpose of data collection, but also for re-use of the data by other migration agencies conducting further analyses; 
  6. establishing formal data sharing arrangements and processes. 

The third step is monitoring any benefits or risks that emerged after data sharing took place

The fourth and final step is evaluating the impacts of sharing, i.e., were the intended benefits achieved and were the risks effectively mitigated, and ensuring that responders have the information they need.

IOM’s Data Sharing Mechanism

In 2007, IOM produced a step-by-step approach to the development of a data sharing mechanism (DSM) for migration data compiled by government agencies (IOM, 2007). This approach, also referred to as the “General Model” for the collection, application and sharing of migration-related data, was piloted in several Eastern European and Central Asian countries. The key premise of the General Model approach is that migration data already exist but are not based on the same concepts and definitions, not aligned with international measurement standards and not being shared by migration actors both within and across countries. The main elements of the General Model are summarized in 6 steps:

  1. Know your counterparts and establish a national network of core institutions and well-defined focal points. A simple operational tool that can be applied during the process of establishing a network is an electronic chart listing the relevant agencies and the contact details of focal points in each agency. Such a chart should include both data producers and data users. There are often problems when staff members are rotated away from their position, which happens regularly in a national network. One possible solution is to assign the responsibility of working within the network not just to a specific individual but also to a specific position, when deciding on a focal point from each relevant agency. In order to ensure that a network remains functional, a national coordinator, or a country administrator (in the DSM terminology), should be nominated.
  2. Working with your counterparts through the establishment of an inter-ministerial working group. There is often only ad hoc or sporadic contact between migration authorities. One way to tackle this is to apply the “whole-of-government approach” by creating an inter-ministerial working group (IMWG). The IMWG does not just make statistics available but also adjusts the available information to different migratory trends over time for policymaking and monitoring purposes. Experience has shown that it is very important that IMWG participants are in a position to take decisions at the appropriate level. Deciding on which authority should be the focal point of the IMWG should be resolved within the country and the solution will differ from one country to another. In some countries, attempting to create an IMWG may end up with no continuation. In most cases, this is an indication that the government has not yet clearly realized the benefits of consolidating migration statistics at the national level and using them for policymaking, assessing, and identifying the common trends of migration and prioritizing available resources. Therefore, much more effort should be put into raising awareness about the importance of statistics and how these numbers can be used for policy making. Another potential reason for the lack of momentum is the absence of a strong data regulatory environment enabling data sharing between agencies. The absence of protocols between national statistical offices and government departments usually prevents sharing of data. The transfer of administrative data to the NSO would be facilitated through a memorandum of understanding (MoU) or other formal agreements covering:

a) the conditions under which data can be used; 

b) the obligations of the statistical agency; 

c) the frequency at which data will be supplied; 

d) the agreed level of detail of the supplied data. 

The government agency might also benefit from the transfer of data to the NSO if the latter provides data analysis services back to the government agency or provides its personnel with training so that they can generate tables for analysis themselves. Administrative agencies may not have the capacity or the expertise to be able to regularly process the data. Therefore, the NSO might assist in the processing of raw data, which can then be transferred back to the relevant agencies for reporting and policy development purposes.

  1. Know your data by mapping out existing infrastructures. The next step is to answer the question “what”? by looking at the migration data itself as it is collected within the country. Developing a meta data file with information on the source of the data collected, the methods used for collecting it and the definitions and meanings behind the terms used. At the end of the mapping out exercise, a chart with an overview of all the statistical indicators and data that are or could be generated in the country can be created.
  2. Know your needs by mapping the demand for migration data at the national level and across agencies. Prioritizing and agreeing upon a minimal set of 5-10 migration indicators based on common definitions and measurement is key to ensuring regular and comprehensive reports.
  3. Know your gaps and use workflow targeted IT upgrades and modern technologies to facilitate the process. IT and workflow assessments look at how agencies and governments manage data at all four stages: collection, storage, aggregation and dissemination.
  4. Enhance your knowledge and skills, for example, through trainings and Train the Trainer courses. An important element in the General Model approach spearheaded by IOM in the European Europe and Central Asia Region is the availability of sufficient expertise and personal capacity in the participating countries to enhance migration data collection and statistics generation. For this purpose, strong emphasis has been made on the need to train migration officials involved in data management: first, the focal points from the agencies involved and then other employees working with the General Model.

Figure 2:

EMDP2C4F2

Case study: observations on the development of a General Model for data sharing in Kazakhstan (IOM, 2007)

The application of IOM’s General Model in Kazakhstan revealed that while data on migration are available, the data are not always being shared, the data are sometimes based on different definitions and the same data are being collected by multiple agencies.  

-In Kazakhstan, the compilation of official statistics is regulated by the 1997 Law on Statistics which designates the National Statistics Agency (ARK) as responsible for the development of a national data strategy. 

-Foreigners planning to stay for more than 6 months are required to register with the Ministry of Interior at the municipality, which also collects information on their country of citizenship, country of birth, date of arrival in country, reason for migration, occupation, education level and marital status. The data is then transferred from the municipality to one of the municipal statistical divisions of the ARK, which checks the data for completeness and accuracy before transferring it to the regional and national statistical divisions. 

-All foreigners entering the country are required to fill a migration card which might be a potential source of data on temporary mobility or ‘movers’, but this data is not being shared with ARK.

-The Ministry of Foreign Affairs also collects data on arrivals and departures of foreigners but the data do not match those of the Ministry of Interior with regards to the definition of an international migrant.

-The Ministry of Justice collects data on citizens who changed their place of residence (based on a 3-month threshold for declaration) which is a potential data source for internal migration, but these data are not shared with ARK.

-Enterprises within the country collect data on foreign employees, which are then transferred to the Ministry of Labour and eventually the Ministry of Interior and ARK.