Why do we need a new dataset?

YOUMIG’s Work Package 4 focuses on the collection of a range of indicators in such policy spheres as population, education, economy and the labour market, and social development. While the YOUMIG dataset is already available online for any interested reader, our other key audience are local decision makers, who are less familiar with working with data. There are a lot of sources of information and data to be found on the internet, so, it may seem that there is no real need for a new dataset. This is not really true. Let me explain.

First, the abundance of information in itself can be a problem. If an official is interested in evidence-based decision making, they are looking to incorporate research evidence into practice in a timely and reliable fashion. This means that for them it is not enough to rely on anecdotal accounts of successful strategies, but must track down evidence, with help of a relevant set of indicators, in order to understand which dimensions should be addressed with policy interventions. The question arises then: what are the indicators they would need to track down? What the YOUMIG researchers have done is to narrow the indicator set in a way that it is highly relevant to certain topics, such as ‘migration’, ‘youth’, or ‘local development’. A datatool – to help visualise and interpret the indicators - is being developed also, to enable decision makers to access the information selected by a multidisciplinary team of experts and partner municipality representatives across the Danube Region. Participant countries are Austria, Bulgaria, Germany, Hungary, Romania, Slovakia, Slovenia, and the Republic of Serbia.

A second reason to collect the YOUMIG dataset is that some important information can be missing or difficult to access otherwise. What we learned, for example, is that many important indicators are only available for the whole population, but not for separate groups, such as age groups and sexes. Similarly, some of the indicators can only be found to represent national averages. Given the differences in the attractiveness of various regions for migration, or the composition of the migration flows across regions and municipalities within a country, we need the indicators also at sub-national level.

 What were the main steps of the data collection and indicator development?

The whole process can be summarised in five main steps.

Experts from IOS Regensburg analysed the goals and recommendations of international bodies  like the European Commission and the United Nations Development Program - in such policy spheres as youth policies, migration and integration policies, and local development policies. The most important issues that populate the political agendas of local and regional institutions were picked out. In addition, the YOUMIG Conceptual Framework and the modified push-pull model developed within it helped us to narrow down the dimensions in which science looks at youth migration. As a result, a very large set of 214 scientifically-grounded indicators were selected, in seven topic areas:

A simple example of an indicator is a country’s GDP, or the number of doctors per thousand people living in a municipality.

The relevance of each indicator was assessed by experts from statistical and research institutions as well as representatives of the YOUMIG partner municipalities. We repeated the selection process, each time excluding indicators that were less relevant to the understanding of youth migration in the context of the participating countries and municipalities. We ended up being able to reduce the set by half.

The remaining 120 indicators were investigated in depth. YOUMIG experts provided a variety of detail about each of them, with the most important being the availability of each and every indicator on a national, regional, and municipality level. They also reported on the degree of desirability of having these indicators included in an imagined evidence-based decision-making scenario, at different levels of governance.

We shortlisted the 16 core indicators with the highest desirability rates across the YOUMIG partnership. These finalists were still scrutinised by the expert teams of statisticians, who refined the definitions and measurement algorithms for each core indicator. All partners assessed the information sources at their hands, and developed detailed plans of the collection process itself, including the exact data sources, the conditions of access to data as well as definitions of which precise data should be requested, and, in some cases, even the formulas of estimation of the indicator values. This was an important step in order to ensure the sustainability of the data collection for those municipalities who would later decide to continue working with these tools that were developed during the two-and-a-half-year lifespan of the YOUMIG project.

Next, we moved from the theoretical considerations and discussions to the down-to-earth issues of collecting values for the core indicators. In this final moment we already wanted to know how, for example, the GDP of a country or the number of doctors in a municipality have changed over the last years. As you may well remember, the guiding principle of the core indicator selection was first of all desirability, and not availability. In such cases, where an indicator was theoretically available and regularly collected by a statistical body in the partner country, our partners have either extracted the information about them from open sources, or made special requests to statistical offices in their countries, in order to be able to access the information needed for YOUMIG’s purposes. In the latter case, some of the information collected is prohibited from sharing with the wider public, while a selected set is made available in the YOUMIG dataset file, which is available online, on the project website.

For the indicators that are not collected by statistical bodies, or in the course of international representative surveys, local partners undertook collection by launching a small-scale survey in their municipality. This data is also only partially open to the public, but can be received on request from YOUMIG data owners, specified in the Data Toolkit under production by the project for each participating municipality.

The newborn YOUMIG dataset is in the end a wonderfully rich data pool, which can help the analysis and better understanding of connections between various social trends affecting youth migration. Besides being the data pool for the YOUMIG Data Toolkit, designed for the general user, it is also an interesting resource for statisticians. They can on the one hand quickly assess whether a desired indicator is available – at national, regional, or municipality level –, and skim over the multitude of values for 16 core indicators across seven participating countries, or, on the other, make judgments about which data they need to procure and how to go about it.[1]

Additionally, by extra effort, some partners carried out surveys locally to supply data for the indicators that were locally important, for example in connection with YOUMIG’s pilot project topics. These indicators were not included in the final core set, but they are available for some YOUMIG partners within the extra indicator set. At the end of this long process, a handbook-style collection of useful indicators was born, as well as a large data file with ready-to-use numbers.

Text by Ekaterina Skoglund, YOUMIG researcher and WP4 leader, IOS Regensburg

Part II of the story coming soon…

[1] Excluding Germany, due to the absence of a participating YOUMIG municipality.