Centre for Economic Demography

Lund University School of Economics and Management

Research School: previous courses

Immigrant Integration
This course examines the demographic and economic impact of immigration from 1945 through 2011 and beyond. Through a study of processes in North America and Europe, we will examine the historical process of immigrant integration into host countries and then extrapolate this process into predictions for the future.

Historical Demography
This course examines sources, methods and theories of historical demography. Themes covered include sources and methods of historical demography (vital registration (parish or civic registration records), family reconstitution, and census-like data), the pre-transitional population system in Europe, the Malthusian system in theory and reality, the Eurasia project on population and family history (comparative analyses of micro-level demography in historical communities in Sweden, Belgium, Japan, Italy and China with a focus on mortality and fertility), and the demographic transition.

Assessing causality by family-based designs
The course starts with a general discussion of causality in epidemiology and the classical validity problems inherent in observational study designs. An overview of common approaches to overcome this caveat in observational epidemiology is given. Family-based studies (e.g., intergenerational, twin and sibling studies) are able to disentangling genetic effects from shared family effects and non-shared individual-level effects. Extensive examples of how family-based studies can be conducted in practice using registry data will be given.

Multilevel Modelling
This course deals with an introduction to multilevel modelling techniques which are quite common in some fields, but quite uncommon in the economic disciplines. The course introduces multilevel data structures and alternative analysis strategies, and deals with subjects such as variance components models, random intercept models, random slope models, contextual effects, three-level models, and modelling variance. Throughout, there is an emphasis on how to interpret the models and on what kinds of research question they can be used to explore.

Sequence Analysis
Sequence analysis is a method which is gaining popularity within the population sciences, especially within those based in sociology. Economic demography has been slow to adopt these techniques, but there exist clear areas in which they should prove quite useful. Based on methods designed to analyze DNA sequences, the use of algorithms to identify spacing between various events was applied to the social sciences in recent years. This course provides an introduction to the concepts involved in sequence analysis and its applications to data of interest to social scientists.

EDSD Courses (2009-2010 & 2010-2011) – These courses are arranged to give first-year PhD candidates the necessary foundations in both demographic theory and quantitative methods which will be needed to produce a high-quality dissertation.

Theories of Demographic Behaviour and Change
The aim of the course is to introduce students to macro-level theories of population change, micro-level theories of demographic behaviour and the micro-macro interactions. At the end of the course, students should comprehend the major theories that explain the level and timing of fertility, family formation and dissolution, the ageing of individuals and society, migration behaviour and migration systems.

Statistical Demography
The course deals with so-called event history models. These are statistical techniques to analyse the occurrence of events in time, such as death, marriage, childbirth, entry into retirement etc. The course covers issues such as characterizing duration distributions and common parametric families, observation schemes (censoring and truncation), nonparametric approaches, basic hazard regression (proportional hazards), the Cox PH model and model diagnostics, discrete-time hazard regression, piece-wise constant hazard model, non-proportional hazards models, and unobserved heterogeneity.

Population Data and Summary Measures
This course provides a detailed study of basic summary measures used in demographic research, and a thorough discussion of both the strengths and the weaknesses of these measures. Data availability and collection are discussed, as are various classification schemes, with an emphasis on causes of death classifications. The course ends with a discussion of how qualitative methods can complement more traditional quantitative methods in demographic research.

Modelling, Simulation, and Forecasting
Modelling, simulations and forecasting are important components of the demographic toolkit. This course teaches techniques needed to analyse the dynamics of age-structured and interacting populations. Through discussions of various indicators the students are taught how to prepare the initial data required for population projections, and then how to forecast population developments based of different scenarios. The course finishes off with an introduction to the fundamentals of microsimulation models.

Consequences of Demographic Change
The course examines the impact of demographic change on the social and economic fabric of society, with a focus on issues of importance to today's policymakers. The course discusses topics such as world population growth, health, social and economic issues related to population ageing, impacts of migration, and gender issues related to a work-family balance. This course has a fairly clear policy focus, with discussion of the wider implications of observed phenomena for national governments and pan-national organizations.

Mathematical Demography
The course covers various advanced topics in formal demography:
- Malthusian and stable populations, Lotka-McKendrick continuous-time and Leslie discrete-time population dynamics
- Matrix population models, techniques of projection
- Cross-sectional vs period index, prevalence-based vs incidence- based index, Markov chains, computation of sojourn times, ageing mechanisms in epidemics
- Tempo effects, two-sex models, inhomogeneous Lotka-Mc Kendrick system
- Frailty models
- Dynamic systems in population and environment, dynamic networks, state-space analysis of differential systems

Causal Inference for Experimental and Observational Studies,
Potential Outcomes, Randomized Experiments and Matching

This course provides a thorough foundation into statistical methods used for causal inference using empirical data. Using the potential outcomes framework of causality, designs and methods for data from randomized experiments and observational studies are discussed. Designs and methods covered in this course module include randomized experiments and matching. Examples are drawn from published research within the economic, medical and social sciences.

The course consisted of several modules dealing with the topic of “Causal Inference for Experimental and Observational Studies”.

Causal Inference Using Potential Outcomes

An introduction to the topic of causal inference. Causal effects will be defined based on the potential outcomes framework of Neyman and Rubin. The fundamental problem of causal inference will be encountered, as well as the discussion on confounding concerning what separates association from causation, and observational studies from randomized experiments. Examples of well-designed observational studies will be introduced, combined with the discussion of the foundations and limitations of statistical models.

Randomized Experiments

The logic of randomized experiments is reviewed, a research design that is widely believed to maximize internal validity and that is becoming ever more popular in the social sciences. Special attention is paid to Fisher's randomization inference, in which randomization is the “sole and reasoned basis for inference". Lastly, we will meet the “Lady tasting tea".

Subclassification and Matching on Covariates

The advantage of randomized experiments is that potential confounders can be safely ignored since they will be balanced, at least in expectations. But randomization is not always practical, nor is it always ethical. How can one ensure valid causal inference in a world without randomization? Designs which assume that selection into the treatment groups is based on observables will be discussed. We start by considering two very intuitive methods, subclassification and exact matching techniques. Next, we discuss matching techniques that are based on the Euclidean distance and the propensity score. We also consider some practical issues with matching such as matching with and without replacement, common support restrictions, and estimating standard errors


Summer course in Historical Demographic Research Using Register-Type Data

This course was organized by the Research School of the Center for Economic Demography (CED), under the auspices of the European Historical Samples Network (EHPS-Net). The course outaly was as follows below and included both theoretical lectures and data management and model estimations by the participants.

  • Vital registration, family reconstitution
  • Scanian Economic Demographic Database
  • Intermediate Data Structure (IDS)
  • Life tables
  • Mortality
  • Marriage and fertility
  • Social stratification and social mobility