Working Paper: CEPR ID: DP16451
Authors: Roel Beetsma; Ward E. Romp
Abstract: Using a new real-time dataset from Beetsma et al. (2020) containing all pension reform measures in 23 OECD countries between 1970 and 2017, we demonstrate that, in contrast to what one might a priori expect, the timing of pension reform measures coincides with business cycle shocks and not with current or projected demographic shocks. We rationalise this finding using a political-economy model with two-sided adjustment costs to explain a lack of response of pension reform measures to changes in demographic indicators.
Keywords: pension reform measures; narrative identification; expansion; contraction; old-age dependency ratio; business cycle indicators; adjustment costs
JEL Codes: H55; H62; J11; J26
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| old-age dependency ratio (J14) | timing of reforms (E69) |
| demographic pressures (J11) | timing of reforms (E69) |
| timing of pension reform measures (H55) | cyclical state of the economy (E32) |
| GDP growth (O49) | likelihood of expansionary reforms (E69) |
| GDP growth (O49) | likelihood of contractionary reforms (E69) |
| unemployment rate (J64) | likelihood of contractionary reforms (E69) |
| public deficit (H62) | likelihood of contractionary reforms (E69) |