Sören edvinsson, Demografiska databasen, Umeå University

2010-2013

The objective of the project In10-0702, POPLINK en världsunik resurs för nydanande forskning om människa och samhälle i ett långt tidsperspektiv, has been to build a multi-generational population database covering Umeå lands- and stadsförsamling, the two largest parishes in the Umeå region, in space as well as in population. Population data from 1900-1950 have been digitized, linked on the individual level, included in a relational database and subsequently being made available for research. The project has been run by the Demographic Data Base at Umeå University, and the database has been built according to established methods using tested and reliable systems with inherent quality controls. No major changes has been made in relation to the original project plan. However, the time plan has been subject to some adjustments, which was mentioned already during the half-time evaluation in 2012. An unexpectedly high number of individual records in both parishes has resulted in a minor delay, and as a consequence the project will not be fully completed until March 2014. As a consequence, the model used for estimation of the number of records in a parish, has been adjusted to better fit the particular context of expanding towns, with generally larger population movements than in most rural areas.

Project results

The particular character of an infrastructural project does at this stage make it difficult to point at specific results, except the fact that we, according to the project plan, have built and provided access to a high-quality multigenerational database with linked population data from Umeå landsförsamling and stadsförsamling, covering the period 1900-1950. The database is constructed for secure and efficient linkage to other regional and national research registers, including the national registers at Statistics Sweden and the National Board of Health and Welfare. Any results proceeding from research on the data, cannot be expected until the first data retrievals have been delivered and analysed, which will at earliest be later this year. One of the first planned projects concerns the fertility decline in Sweden on an individual and a family level, throughout the first half of the 20th century.

In all, the project has followed the plan which was set out in the application to the Swedish Foundation for Humanities and Social Sciences. Population data on the individual level in Umeå landsförsamling and Umeå stadsförsamling for the period 1900-1950, including longitudinal population registers, birth records, marriage records, burial records and migration records, have been digitized into a database following consistent and established principles. Data entry is succeeded by a number of systematic linkage procedures, according to carefully stipulated criteria. Individual records from different sources are linked together into life biographies, family and kinship is defined and parishes are linked together, forming larger geographical regions and making it possible to trace and follow individuals and families between different parishes in the database. This makes the population database well suited to life-course- and multigenerational studies and useful for micro-studies on the individual and community level as well as for large-scale projects requiring large amounts of data. As a final step of the data production process, a separate key table for safe and secure linkage to other registers is created and delivered to Statistics Sweden.

The project activities have been planned and executed according to a detailed scheme with current quality controls and monthly checkpoints. The follow-up includes a number of standard measures such as process time, capacity and progress in relation to the project plan and monthly feedback is given to the staff. Between 2011 and 2013 the total project time counts to 52 640 hours, of which 32 588 hours have been accounted for during work with Umeå landsförsamling and 20 052 for work with Umeå stadsförsamling. In Umeå 278 562 individual records have been digitized and linked and in Umeå stadsförsamling 214 615 records have been digitized. The linkage is in progress, and will, slightly delayed, be completed in March 2014. Umeå landsförsamling was completed during quarter 3 in 2013 and both parishes will be ready for research by quarter 2 in 2014. The project economy has followed the plan, which was drawn up at the onset of the project. The project grant has been used to finance part of the project; the remaining resources have consisted of co-financing by the Demographic Data Base.

Unexpected problems

As mentioned earlier, the number of records in both parishes was underestimated in the model that was used when the first project plan was established. As a consequence the project is slightly behind time in comparison to what was expected at the onset of the project.

Integration of work in the organization and future developments

The actual project is part of a larger incentive to build a new multigenerational research infrastructure for longitudinal studies within social sciences, humanities and life sciences, covering large parts of the population in Västerbotten until 1950. When linked with today's administrative research registers, it will produce a unique and powerful resource of international relevance, which for the first time makes it possible to take advantage of the full scientific potential of the rich Swedish registry data for cutting edge research. The focus on the population of Västerbotten creates valuable synergies with several unique population-based registers for the region. It opens for novel research initiatives within a large number of scientific fields, where a multigenerational approach can contribute with new perspectives and insights. It will for example be possible to conduct in-depth studies of the large-scale transformations affecting 20th century Sweden in a longitudinal perspective, and thus achieve a better understanding of the finer mechanisms behind social and demographic change affecting households, labour markets, welfare systems and public health.
A long-term plan has been drawn up for the future developments and potential sources for funding are consecutively sought for.

New research questions generated in the project

An infrastructural project of this character, aiming at creating new and improved opportunities for future research rather than producing research, do not generate new research questions in the same way as conventional research projects regularly do. However, this does not mean that building and establishing a general infrastructure - in this case, a research database - is a totally predictable process, which always runs smoothly. Also in infrastructural projects questions arise which needs to be addressed, questions which affect and frequently also develop the database building process. There are numerous inevitable source-related problems to consider, which needs to be handled in a unified way. These often concern issues related to the application of present principles in new situations, but frequently also involves re-thinking and re-interpretation of old principles, for example when new family forms become more common in the sources. The entire data production process, including the issues and questions that need to be addressed and handled throughout a project, is carefully documented according to established routines. It shall be possible to control how the information in the sources is transcribed and coded into the database, down to almost every single record. Throughout the work with the Umeå parishes we have had to direct particular attention to refining and developing our methods and principles for handling residential information in the context of an expanding and rapidly developing town.

A distinct spin-off effect of this project is the increasing interest for 20th century population data that we have met from researchers, nationally as well as internationally. Several contacts have been established, of which some have resulted in tangible project plans and in grant applications, based on data in POPLINK.