The third pillar in Europe: institutional factors and individual decisions
Accompanied by pension reforms, most European countries have introduced taxdeferred individual retirement accounts as a means to incentivise private, voluntary savings for retirement in the “third pillar”. The introduction of these accounts has opened chances and risks for their owners: on the one hand, households can decide whether to save additionally for retirement and are rewarded with substantial taxdeferrals, on the other hand they may lack the financial knowledge to save voluntarily in these schemse and be left with insufficient retirement savings.
This paper focuses on how the characteristics of households in different countries are correlated with voluntary retirement saving. We use the most comprehensive European data set on portfolios of households aged 50 and above currently available which provides us with fully comparable data across 11 European countries. We document and study differences and similarities in ownership of tax-deferred retirement accounts. Descriptive statistics reveal that households in northern European countries where other types of pre-funded individual accounts exist in the public pension system have the highest ownership rates. Given the countries’ different pension systems, we then calculate the effects that households’ characteristics in each country have on the probability to own IRAs. We find that among households in countries which experienced structural retirement reforms educational differences matter less than in those countries where the introduction of IRAs is relatively recent and where pension systems are relatively generous. Additionally, other variables related to knowledge spill overs such as social activities and cognitive functions matter. Among the working population in our sample, educational differences play a major role for and explain differences in ownership up to 20%. Pooling households from all countries together, we find that the probability to hold an IRA is any country is substantially lower than in Sweden which we attribute to the institutional environment in this country.