Panel Data Analysis

Author

Affiliation

UCL Social Research Institute

Introduction

With usual cross-sectional data, we only observe each unit once.

With panel data, we observe the same unit (person, region, or country) repeatedly over time.

This allows applying statistical methods which identify relations based on the within-unit changes rather than the differences between units. We can then account for some statistical problems of non-experimental studies and increases the confidence in a causal interpretation of the results.

This course provides a hands on introduction to the preparation of panel data and the application of panel data methods. The course focuses on the most common methods like Difference in Differences, Random Effects, and Fixed Effects. We will also briefly discuss some extensions like novel Diff-in-Diff estimators and Fixed Effects Individual Slopes. The empirical part will be based on R and RMarkdown.

This course profited a lot from teaching materials by Josef Brüderl and Volker Ludwig: Slides

Why do we need panel data analysis?

In empirical social sciences, we are often interested in the causal research questions: we want to investigate questions of cause and effect. Using purely cross-sectional data, we can compare two (or more) different units (e.g. people, states, companies) and test is there exist differences according to our treatment or variable of interest.

However, these units may not only differ in the dimension of interest, but may also be different in many other aspects, and it is likely that we do not obverse all the relevant aspects. In the end, we might erroneously conclude that there is a causal effect, even though we only observe correlation (which is confounded by other differences).

Randomized controlled trials (RCT) provide a way to circumvent the problems of unobserved differences between treatment and control cases.

By randomly selecting some individuals and exposing them to the treatment of interest, we make sure that no (unobserved) characteristics of these individual can be correlated with the treatment. In other words, units in the treatment group should - on average - be identical to units in the control group on all characteristics except the treatment.

However, randomly exposing some individuals to treatment and withholding treatment from others can be tricky in the social sciences. Think about the effects of education, the effects of marriage, or the effect of pregnancy (good luck with your ethics committee).

A potential middle ground between those two approaches: “compare alike with alike” (Firebaugh 2008). This is what we usually aim for by using panel data. We do not compare two different units to each other. Rather, we compare a unit in an earlier stage to the same unit in a later stage. Not as save as an RCT, but much more ethical for many research questions.

Illustration by Huntington-Klein (2021): Introducing a binary control variable (left) and introducing fixed effects (right)

Some examples

Plant openings and housing prices (Currie et al. 2015)

Life course events and happiness (Clark and Georgellis 2013)

London Congestion Charge and School Attendance (Conte Keivabu and Rüttenauer 2022)

Extreme Weather Events Elevate Climate Change Belief but not Pro-Environmental Behaviour (Rüttenauer 2023)

There are obviously plenty of other examples on various topics, such the determinants of life satisfaction (Gattig and Minkus 2021; Kapelle et al. 2022), family dynamics and the gender wage gap Zoch (2023), or drivers of volunteering and pro-social behaviour (Aksoy and Wiertz 2023; Dederichs and Kruse 2023). Moreover, panel data and Differences-in-Differences designs occupy a prominent position the evaluation of social policy impacts (Conte Keivabu 2022; Franzen and Bahr 2023; Goodair and Reeves 2022; Mader and Rüttenauer 2022; Kneip and Bauer 2009).

Further materials

Extensive slides by Josef Brüderl and Volker Ludwig: Slides

References

Aksoy, Ozan, and Dingeman Wiertz. 2023. “The Impact of Religious Involvement on Trust, Volunteering, and Perceived Cooperativeness: Evidence from Two British Panels.” European Sociological Review, April, jcad024. https://doi.org/10.1093/esr/jcad024.

Allison, Paul David. 2009. Fixed Effects Regression Models. Vol. 160. Quantitative Applications in the Social Sciences. Los Angeles: Sage.

Angrist, Joshua David, and Jörn-Steffen Pischke. 2015. Mastering ’Metrics: The Path from Cause to Effect. Princeton: Princeton Univ. Press.

Brüderl, Josef, and Volker Ludwig. 2015. “Fixed-Effects Panel Regression.” In The Sage Handbook of Regression Analysis and Causal Inference, edited by Henning Best and Christof Wolf, 327–57. Los Angeles: Sage.

Clark, Andrew E., and Yannis Georgellis. 2013. “Back to Baseline in Britain: Adaptation in the British Household Panel Survey.” Economica 80 (319): 496–512. https://doi.org/10.1111/ecca.12007.

Conte Keivabu, Risto. 2022. “Extreme Temperature and Mortality by Educational Attainment in Spain, 2012.” European Journal of Population 38 (5): 1145–82. https://doi.org/10.1007/s10680-022-09641-4.

Conte Keivabu, Risto, and Tobias Rüttenauer. 2022. “London Congestion Charge: The Impact on Air Pollution and School Attendance by Socioeconomic Status.” Population and Environment 43 (4): 576–96. https://doi.org/10.1007/s11111-022-00401-4.

Croissant, Yves, and Giovanni Millo. 2019. Panel Data Econometrics with R. Hoboken, NJ: John Wiley and Sons.

Cunningham, Scott. 2021. Causal Inference: The Mixtape. New Haven and London: Yale University Press.

Currie, Janet, Lucas Davis, Michael Greenstone, and Reed Walker. 2015. “Environmental Health Risks and Housing Values: Evidence from 1,600 Toxic Plant Openings and Closings.” American Economic Review 105 (2): 678–709. https://doi.org/10.1257/aer.20121656.

Dederichs, Kasimir, and Hanno Kruse. 2023. “Who Stays Involved? A Longitudinal Study on Adolescents’ Participation in Voluntary Associations in Germany.” European Sociological Review 39 (1): 30–43. https://doi.org/10.1093/esr/jcac013.

Firebaugh, Glenn. 2008. Seven Rules for Social Research. Princeton, N.J. and Woodstock: Princeton University Press.

Franzen, Axel, and Sebastian Bahr. 2023. “Poverty in Europe: How Long-term Poverty Developed Following the Financial Crisis and What Drives It.” International Journal of Social Welfare, May, ijsw.12614. https://doi.org/10.1111/ijsw.12614.

Gattig, Alexander, and Lara Minkus. 2021. “Does Marriage Increase Couples’ Life Satisfaction?: Evidence Using Panel Data and Fixed-effects Individual Slopes.” Comparative Population Studies 46 (May). https://doi.org/10.12765/CPoS-2021-05.

Goodair, Benjamin, and Aaron Reeves. 2022. “Outsourcing Health-Care Services to the Private Sector and Treatable Mortality Rates in England, 2013: An Observational Study of NHS Privatisation.” The Lancet Public Health 7 (7): e638–46. https://doi.org/10.1016/S2468-2667(22)00133-5.

Huntington-Klein, Nick. 2021. The Effect: An Introduction to Research Design and Causality. Boca Raton: Chapman & Hall/CRC.

Kapelle, Nicole, Theresa Nutz, Daria Tisch, Manuel Schechtl, Philipp M. Lersch, and Emanuela Struffolino. 2022. “My Wealth, (Y)Our Life Satisfaction? Sole and Joint Wealth Ownership and Life Satisfaction in Marriage.” European Journal of Population 38 (4): 811–34. https://doi.org/10.1007/s10680-022-09630-7.

Kneip, Thorsten, and Gerrit Bauer. 2009. “Did Unilateral Divorce Laws Raise Divorce Rates in Western Europe?” Journal of Marriage and the Family 71 (3): 592–607. https://doi.org/10.1111/j.1741-3737.2009.00621.x.

Mader, Sebastian, and Tobias Rüttenauer. 2022. “The Effects of Non-pharmaceutical Interventions on COVID-19 Mortality: A Generalized Synthetic Control Approach Across 169 Countries.” Frontiers in Public Health 10: 1–8. https://doi.org/10.3389/fpubh.2022.820642.

Rüttenauer, Tobias. 2023. “More Talk, No Action? The Link Between Exposure to Extreme Weather Events, Climate Change Belief and Pro-Environmental Behaviour.” European Societies, November, 1–25. https://doi.org/10.1080/14616696.2023.2277281.

Schmitt, Laila, and Katrin Auspurg. 2022. “A Stall Only on the Surface? Working Hours and the Persistence of the Gender Wage Gap in Western Germany 1985.” European Sociological Review 38 (5): 754–69. https://doi.org/10.1093/esr/jcac001.

Van Winkle, Zachary, and Anette Eva Fasang. 2020. “Parenthood Wage Gaps Across the Life Course: A Comparison by Gender and Race.” Journal of Marriage and Family 82 (5): 1515–33. https://doi.org/10.1111/jomf.12713.

Wooldridge, Jeffrey M. 2010. Econometric Analysis of Cross Section and Panel Data. Cambridge, Mass.: MIT Press.

Zoch, Gundula. 2023. “Participation in Job-Related Training: Is There a Parenthood Training Penalty?” Work, Employment and Society 37 (1): 274–92. https://doi.org/10.1177/09500170221128692.