- 15 May 2015
Italy's pension reforms: facing the factsNicholas Barr* and Peter Diamond**
Italy’s pension reforms have made the pension system simpler, fairer, more sustainable and more transparent. They are a strategy on which to build. They should not be undone. The whole issue has been brought into sharp focus by last Thursday’s Constitutional Court decision that the temporary removal of inflation adjustment for pensions was unconstitutional.
Though short-term economic pressures were the trigger for urgent action, the underlying problem is long-term, and the response addressed both sets of issues. Italy is ageing rapidly, with a fertility rate well below the OECD average and life expectancy among the highest in the OECD. As a result, the number of people of working age per person above 65 (known as the old-age dependency ratio) has fallen from 4.2 in 1980 to 2.8 in 2014, and is projected to fall to 1.5 in 2050.
But there is also a short-term problem. In 2011, public pension spending was 9% of GDP, almost the highest in the OECD (only Greece and Spain spent more), and 1¾ times the OECD average of 5.2%. The comparable figures for Germany and Sweden were 6.9% and 6.2%, respectively (OECD Pensions at a Glance 2013, Table 6.1).
In part, this high cost reflects relatively early retirements. Italy’s effective retirement age is one of the lowest in the OECD. And labor market participation of older workers (55-64) remains relatively low despite an increase from 27.7% in 2000 to 40.4% in 2012.
Reform by the Monti technocratic government aimed at bringing public spending generally and pension spending in particular closer to the OECD average.
The changes laid down a comprehensive reform. The main measures included:
(1) accelerating the reforms legislated in the 1990s, whose implementation had been slow for political reasons, (2) raising the retirement age, (3) moving towards equalising retirement ages for men and women, and (4) improving the mechanism for adjustments in the future.
Two questions arise: was there any choice about the overall direction of reform; and were the reforms good ones?
In principle, the policy makers had three and only three responses available to be combined to address the falling old-age dependency ratio. They could:
1) Increase pension spending to help cover the extra cost. If that increase is financed entirely by contributions, the higher pension spending is financed by workers.
2) Reduce the average monthly benefit to help keep pension spending from growing so much. In that case, the cost of adjustment falls on future pensioners’ living standards in retirement.
3) Increase pension age (with no corresponding increase in pensions) to help keep spending from growing so much. In this case, the cost of adjustment falls on future pensioners, not through lower living standards in retirement but through a shorter duration of retirement.
Good policy design should have two characteristics. First, it is likely to use more than one of these instruments, and possibly all three. Second, it will phase in changes gradually and predictably; pensions should assist long-run planning, so sharp short-run shocks are undesirable.
If politicians ignore predictable (and predicted) future problems, two undesirable outcomes are likely to occur.
Pension spending will increase by more than is desirable.
Eventually a crisis will force a sudden reform.
The combined effects of the 2008 economic crisis and the eurozone crisis, coming on top of years of delayed responses to the pension problem, was a stress test too far for Italy, requiring reform that was quicker and deeper than would otherwise have been desirable.
High pension spending came from generous access to benefits, inadequate adjustment for rising life expectancy and incomplete reforms. Accompanying the high spending were almost the highest contribution rates in the OECD (33% in 2012, compared with an OECD average of 19.6%; OECD Pensions at a Glance 2013, Table 6.1).
This legacy of many years of a large problem being left for future governments was the inheritance of Elsa Fornero, the Minister of Labor, Social Policies and Gender Equality in the Monti government from November 2011 to April 2013.
The main Monti-Fornero reforms included all three of the responses above.
The existing high contribution rates mostly ruled out Response 1, the exception being higher contribution rates for self-employed workers. So, the problem had to be addressed mostly by a combination of Responses 2 (lower monthly benefits) and 3 (later access to benefits). It is important to be clear that there are no other possible responses.
Response 2 included policies that reduced monthly benefits by speeding up previously-legislated reform, together with automatic adjustment of benefit levels for life expectancy, thus relating benefits more closely to a person’s contribution record.
Response 3, to delay the start to benefits included:
increasing retirement age to 66 in 2012 for men and for women in the public sector, and for all women from 2018.
Linking normal pension age and life expectancy from 2013, with automatic updates subsequently.
It is widely asserted that later retirement aggravates youth unemployment. Both theory and international evidence show that that argument is wrong. If it were true, countries with later retirement would have higher rates of youth unemployment. They do not.
A pension reform that needs to be implemented relatively rapidly always faces delicate transition problems. Here, the situation of the esodati – workers who had left the labor force early expecting to be able to draw pensions soon – was a particular problem. The INPS told the government that the number of workers involved would be small.
Faced with a large number of esodati (estimates vary between 65,000 and 350,000), the Monti government found a solution for some 130,000, and legislation in 2014 brought in safeguards for further numbers. With hindsight, the government should have moved faster once the true scale of the problem became known. That said, going forward, any adjustment to address this problem should not undermine hard-won sustainability.
While some of the reforms had little direct effects on pension spending, importantly, they improved the pension system in other ways, in particular its links with the labor market, including greater simplification and transparency; better-designed flexibility for workers over when to retire; and gender harmonisation of rules.
At a strategic level, these reforms were both right and necessary. They were right in that they brought pensions in Italy closer to best practice internationally – desirable reform directions even in the absence of a fiscal crisis.
Though Italy introduced pioneering reform earlier, it had not gone far enough or fast enough. Reform was also necessary for fiscal reasons. In an ideal world, the reforms would have been phased in gradually over the preceding decades. But by 2011 the problem had become too large and too urgent for gradual reform.
There is an inherent tension between the long-term time horizon of good pension design and the short-term pressures of politics. The Fornero reforms put Italy’s long-run interests ahead of short-term political advantage. Thus they were not only right; they were brave.
* Professor of Public Economics, London School of Economics
** Institute Professor Emeritus, Massachusetts Institute of Technology and Nobel Laureate in Economics