Tomáš Havránek, Marek Rusnák, Anna Sokolova
We examine 567 estimates of habit formation from 69 studies published in peer-reviewed journals. In contrast to previous results for most fields of empirical economics, we find no publication bias in the literature. The median estimated strength of habit formation equals 0.4, but the estimates vary widely both within and across studies. We use Bayesian model averaging to assign a pattern to this variance while taking into account model uncertainty. Studies using micro data report consistently smaller estimates than macro studies: 0.1 vs. 0.6 on average. The difference remains large when we control for 21 other study aspects, such as data frequency, geographical coverage, variable definition, estimation approach, and publication characteristics. We also find that estimates of external habit formation tend to be substantially larger than those of internal habits, that evidence for habits weakens when researchers use higher data frequencies, and that estimates differ systematically across countries.
JEL codes: C83, D12, E21
Keywords: Bayesian model averaging, consumption, habit formation, meta-analysis, publication bias
Issued: May 2015
Download: CNB WP No. 3/2015 (pdf, 539 kB)