Reflecting on an internship with the Department of Work and Pensions

Author: Clara Girault, University of Southampton,

Bio:

Clara Girault is a third year PhD student at the University of Southampton in Social Statistics and Demography and funded by the ESRC South Coast Doctoral Training Partnership. Her research focuses on the estimation of hidden suicide within events of undetermined intent in high-income countries’ official statistics, and more specifically on predictive methods using multiple causes-of-death and socio-demographic information found in death certificates.

From September to December 2024, I underwent an internship at the Department of Work and Pensions (DWP) in the Demography Centre of Expertise. The internship focused on updating the partnership matching assumptions for the DWP’s dynamic microsimulation model of pension accrual and decumulation (Pensim).

Why update the assumptions on partnership matching and separation? Firstly, the initial assumption used in the projection did not consider same-sex couples. Secondly, the previous assumptions were based on older data, and we know that couples nowadays do not have the same profiles of partnership (who enters a relationship with whom), nor do they have the same rates and profiles at the end of the partnership (who ends a relationship with whom). The type of relationship also changed, with more cohabitant couples out of marriage or civil union.

We used two waves of Understanding Society to answer the following research questions:

• Which characteristics most influence whether a person will form a partnership in the next year?
• What is the probability of someone with those characteristics forming a partnership in the next year?
• Which characteristics most influence who a person chooses as their partner?

Unfortunately, we could not test those assumptions in the model to see how the outcomes evolve since the update. Due to the small sample of same-sex couples and the coding issue of individuals evolving from same- to different- (and vice-versa) sex couples, this part of the analysis consisted solely of the code and how to identify those couples. Understanding Society includes a file in which individuals state their relationship with others in the household. By using this method to identify couples, we could study outside of legal unions and consider the rising number of cohabitant relationships.

At the DWP, analyses are mostly made using the programming language SAS, which I haven’t used in quite a while Hence, an important share of my 3-month internship focused on translating SAS to R code, and back again to SAS, as most government statistical institutes now move to R. It was a challenge, but I have learned a lot and gained valuable analytical skills by doing so.

The internship ended with a presentation within the Pensions and Later Life Analysis (PALLA) team.

Overall, this internship was extremely interesting. Not only by its research aspect (partnership matching and dissolution) but also for its governmental environment. I discovered how a demographer in a government department works and how our research is used for policies. Although I knew that demography is a major aspect in political decision-making, I was unaware of how assumption updates are carried out. Besides the work environment, my research has always focused on mortality studies and mental health using demographic methods. It was refreshing to find an internship that did not focus on my PhD subject, and to know that the skills I have learned could be applied to completely different subjects. In September 2025, a poster was presented by the Demography Centre of Expertise at the British Society of Population Studies (BSPS) annual conference, which included some of the work I was doing while at DWP.

I am glad that I took the opportunity of doing this PhD placement, and I would encourage other PGR students to do so. Having a “pause” in the PhD research project can be genuinely challenging, but the potential of learning new methods, discovering a new working environment, and undergoing an internship outside of the academic research setting is, I believe, a valuable addition for young researchers.