Abstract
Identifying the causes of happiness presents a challenge for researchers interested in this fundamental outcome variable. After reviewing previous literature looking at the causal effect of political participation on life satisfaction, we discuss the merits of using panel data, where there are repeated measurements over time for each individual, and discuss two common statistical models used in the analysis of panel data, the autoregressive distributed lag model, and the fixed effects model. We use both techniques to analyze the British Household Panel Survey and find evidence that social participation strongly predicts life satisfaction but not that voting participation predicts life satisfaction. We argue that the panel data models help reduce the risk of time-invariant omitted variable bias but are still subject to the problems of time-varying omitted variables and reverse causality. The article aims to provide guidance to researchers seeking to analyze the determinants of life satisfaction using large survey data sets.
Original language | English (US) |
---|---|
Pages (from-to) | 5-23 |
Number of pages | 19 |
Journal | Voluntas |
Volume | 26 |
Issue number | 1 |
DOIs | |
State | Published - Feb 2014 |
Keywords
- Happiness
- Panel data analysis
- Political participation
- Survey data
- Voting
ASJC Scopus subject areas
- Business and International Management
- Sociology and Political Science
- Public Administration
- Strategy and Management