PensionReforms summary and comments
This 2003, 12 country report from the World Bank paints a less than encouraging picture of Latin America's compulsory Tier 2 arrangements (see also here for another similar review from 2006).  The first paragraph of this 110 page piece sets the tone:

"Some areas of public policy require very long-term horizons. Global warming and pension policies are two good examples. In these areas, decisions today produce consequences many decades into the future yet there is inevitably a high degree of uncertainty as to the outcome. For this reason, the original course plotted must be continuously corrected in light of new evidence. No pension system that exists today is likely to remain exactly the same by the time young affiliates retire in forty years."

The report suggests six elements that drove reform in Latin America, starting with Chile in 1981:
"1. Increasing pressure on the central budget to cover deficits
2. Inequities within and across schemes
3. Lack of long term sustainability due to internal demographics
4. Failure to provide promised benefits
5. Poor management of reserves, past or present
6. Distortion of labor market and incentives to evade."

Changes were undoubtedly needed - poor governance, evasion and "and erosion of credibility" in the old systems meant that something had to happen.  The report suggests that the failure to deliver past promises was probably the most important driver, rather than fiscal pressure, increasing saving, developing capital markets or privatisation.

"The new model of pension provision has offered a viable alternative to the failing pay-as-you-go, public schemes that had become prevalent around the world over the last century.  In its short history in Latin America, comprising 55 years of cumulative experience, millions of workers have opened accounts in specialized firms under strict supervision, acquired property rights and a greater degree of control over their retirement income.  In general, they have enjoyed good rates of return, more flexibility in terms of benefits and better service in the new system."

The report describes the Latin American experience as being "at a cross roads":
"Despite the positive early experience, .. only Chile can be said to have consolidated its reform and to have taken advantage of the ancillary benefits to its economy.  There is little doubt as to the credibility of the system, its robustness and its role in providing retirement income for Chilean workers.  Regarding the impact on the economy, Chile's experience suggests that under the right conditions, the benefits of the new model can extend beyond affiliates of the system through various channels that increase economic growth."

In the other 11 countries covered, the reforms ". are not consolidated and in a few countries, proposals to reverse reforms are part of the public debate.  A more insidious threat however, are the more frequent proposals that undermine the basic philosophy of the reform in an attempt to channel funds to a politically expedient purpose.  This myopia has already gone some way towards compromising a key feature of the system, namely, transparency of individual account values in El Salvador and Bolivia.  A somewhat different case is that of Argentina where it was proven that property rights are never fully protected.  While individual accounts may be better insulated, they are no more immune from expropriation than bank deposits."

Those are not the only challenges to the new framework.
"Small countries with low incomes and coverage face extra challenges.  The nature of the industry combined with the perceived need to impose onerous regulatory burdens during the initial years of operation leads to highly concentrated markets.  Supervision costs are higher.  Capital markets and insurance sectors are weak.  A concerted effort involving multilateral and bilateral partners to tackle similar problems facing the countries with fewer than one million affiliates would be useful at this stage."

More strenuous efforts are needed to shore up support for the Tier 2 arrangements:
"The most effective defenders of the basic principles and sound running of the new pension systems are well-informed affiliates and independent supervisors.  Long run self-preservation also makes the pension fund industry a key ally, (albeit with selfish interests that lead to the need for effective supervision).  Academia, nongovernmental organizations and international organizations can all support efforts to improve the system.  The stakeholders face a major challenge to ensure the viability and success of the new systems."

The report lists what the measures of "success" might be when looking backwards from a point 30 years hence:
". The explicit objectives of the system - including overall benefit targets, acceptable dispersion of outcomes, balance of private DC and public DB, and funding targets are understood by policymakers and the public at large
. Minimum pension guarantees are linked to objective indices based on clear poverty objectives or replaced by social assistance schemes with greater coverage
. Unfunded liabilities due to residual public DB schemes or minimum pension guarantees are affordable
. All covered workers participate in the funded scheme
. Overall coverage of the labor force has increased in line with income per capita
. On average, accounts have earned a net return-wage differential of two percentage points or more over any 20 year sub-period
. The charge ratio for full career worker is below 15 percent
. Property rights have [not] been infringed upon through expropriation
. There have been no failures of pension funds or large government bailouts
. Significant diversification of country-specific risks
. Significant diversification into private domestic securities
. Most workers have a reasonable understanding of their provider and portfolio choices and are able to compare prices
. Supervisors are institutionally and financially independent and are not beholden to the executive branch of government
. The sum of the implicit pension debt and the explicit conventional debt is less than what was projected prior to the reform
. A majority of the voting population would reject proposals to revert back to the old pension model".

PensionReforms suggests that the report is a less than ringing endorsement of even the 20 year Chilean experiment.  There are limited exemptions from coverage in Chile but those do not come close to explaining why only 34% of the Chilean working age population were contributors in 2002.  An unweighted average of 10 countries has the average membership at only 23% of the working age population.  In none does it exceed 50%.  The report notes that these percentages are of the whole working age population rather than the more usual "economically active".  "Demographic indicators avoid measurement and definitional problems but do not make it clear what proportion of the potential universe of contributors the schemes cover."

The consolidation of providers indicates that they have struggled to make the business pay, probably due to the controlled, artificial markets that are needed to protect the public policy reasons for having the schemes.  Chile started with 13 providers (AFPs) - that grew to 20 by 1992 but now (2002) stands at only seven, three of which have more than 70% of contributors.

Change was undoubtedly required in the then existing pension systems that were simply not delivering.  The question not addressed in this report is whether the changes that happened were the most likely to succeed.  Turning to the six elements that the report suggests drove the reforms, there is no doubt that fiscal deficits needed to be addressed; inequities undermined the credibility of existing arrangements; the Defined Benefit schemes needed to deliver on their "promises" and, because they didn't, were therefore unsustainable.  Finally, there were significant failures of governance.

Compulsory Tier 2 arrangements do not necessarily fix those difficulties, as the report effectively acknowledges.  They change the nature of the issues but are themselves subject to the same kinds of pressures as there is so much money and such a long time involved before the new system might be described as 'mature'.  Equity, credibility, and good governance are just as important in the new environment.  Their consequences are perhaps easier to see in a Defined Contribution environment but are really no easier to control.  That's why compulsory Tier 2 schemes need so many rules.  (File size 334 KB)  212