Kazakhstan: An Ambitious Approach to Pension Reform
By Emily S. Andrews (ECSHD, World Bank).
A far-reaching and bold Kazakhstan pension reform law was enacted in June 1997 and inaugurated in January 1998. The reform immediately transformed the pension system from an expensive pay-as-you-go (PAYG) system to one of fully funded, defined contribution accounts. The previous system suffered from three overarching problems: (i) a soviet-style structure of benefits characterized by early retirement ages, special privileges, and high replacement rates, particularly for short-service workers; (ii) excessively high payroll taxes for pensions totaling 25.5 percent of the wage bill, and (iii) a continually declining revenue base which resulted in an estimated two-thirds of projected revenue collected from state owned enterprises and corporate employers and, at most, an estimated one percent of revenue from small business in the private sector. In addition, all attempts at a more restricted reform of the pay-as-you-go (PAYGO) system had essentially ended in failure – that is without an increase in revenues or a reduction in cost.
The new system, a funded system based on investments, is intended to promote self-sufficiency instead of government dependence, to reduce government expenditures, to encourage savings and promote the development of the capital market. Starting January 1, 1998, the pension system was transformed into one of fully funded, individual defined contribution accounts mandated for all workers on an immediate basis. Accrued entitlements to old pensions were maintained, however, so that upon pensionable age workers would receive pensions for service prior to January 1, 1998 from the old system and benefits for service after January 1, 1998 from their contribution accounts.
Current PAYGO pensions and future partial PAYGO pensions initially were to be financed from a 15-percent wage tax paid by employers, a reduction from the 1997 25.5-percent PAYGO contribution rate. This form of financing was selected rather than the recognition bond approach instituted under the Chilean system. Over time, as obligations under the old system would be phased-out, the 15 percent payroll tax was to be phased-out as well. Starting January 1, 1999, however, payroll taxes earmarked for social programs (pensions, unemployment insurance, and health insurance) were amalgamated into a single 21-percent payroll tax included in the budget as another source of general revenues. Presumably, when the PAYGO expenditures decline in the future, the overall tax burden can be reduced.
Pensions under the new system are based on contributions of 10 percent of earnings allocated to individual accounts and invested in financial instruments in accumulation funds. Competing accumulation funds, including one managed by the state, have been established. The pension reform legislation is underpinned by complementary safety-net legislation which provides social allowances for disability, old-age, and the loss of a wage-earner.
Government realizes the need to inform the public about such a substantial change in their pension system and has responded, in part, by a nationwide effort to put key reform officials before the public, in person and through the media, to provide information about the reform at the highest level. Continued concentrated public information efforts still are necessary to win public confidence and to ensure that mandatory pension contributions are made and wage taxes paid. Government has prepared a comprehensive plan for such a public information campaign, with a timetable for the implementation of the public information strategy. That strategy will be supported by technical assistance from the Asian Development Bank (ADB) and the World Bank.
The pension reform program is not without fiscal costs. When the reform was initiated in 1997, these costs conformed to a sustainable fiscal policy objective. The incremental cost of pension reform in 1998 was estimated at 1.7 percent of GDP or US$395 million. Government programmed an adjustment in the budget of 0.7 percent of GDP for 1998 to offset the cost of pension reform. Initially, the loss of revenue was to be almost entirely made up by pension fund investments in government bonds. In 1998, however, payroll tax revenues were only around two-thirds of the revenues collected in 1997. Yet this shortfall was not primarily a result of the reform itself. Through the reduction of the income tax, the tax base upon which the payroll tax was collected was smaller. Further, the responsibility for the collection of the 15-percent payroll tax had been shifted from the oblasts to the tax administration. As a result, the local authorities had little remaining incentive to encourage enterprises to pay. In the longer run, the centralization of tax collections, combined with current initiatives being developed to improve tax administration, should increase revenues substantially.
The impact of the economic downturn resulting from the joint effects of the Asian crisis, the Russian collapse, and the worldwide decline in basic commodity prices, especially oil, have reduced the revenues coming to Government. Nonetheless, Government has taken appropriate decisions to stabilize the economy through fiscal and monetary policy measures. The future of the pension reform will largely be tied to continued macroeconomic stability. While current circumstances make it more difficult to sustain the reform, the growth of a burgeoning, vibrant new private pension industry calls for renewed support, particularly under such difficult economic circumstances.
This paper reviews the events which led to the reform by describing in some detail the failings of the former system. Then the new institutions resulting from the reform are described – the pension plans, the asset management companies (AMCs), and the custodian banks. The paper goes on to describe the structure, functions, and progress of the three new regulatory offices – the National Pension Agency (NPA), the regulatory office of the National Securities Commission (NSC), and the regulatory office of the National Bank of Kazakhstan (NBK). Following that, the benefit structure is reviewed and requirements for the future developed. The initial analysis of the fiscal sustainability of the reform is reviewed and future sustainability discussed in the context of the recent exogenous economic shocks. Finally, the paper explains the decision of the World Bank to support the Kazakhstan pension reform through a US$300 million structural adjustment loan and a smaller US$17.4 million implementation project. The analysis concludes that: (i) the ambitious reform was inevitable; (ii) the reform is progressing in a satisfactory manner despite external circumstances that have restricted the prospect for strong economic growth and rapid capital market development; and (iii) the reform was and is worthy of World Bank support.
II. The Need for An Ambitious Pension Reform
Government decided on an ambitious pension reform primarily as a result of serious arrears in pension payments in an environment that made a more moderate reform unlikely. The problems under the old system were manifold. First, retirement ages were extremely low, at age 55 for women and age 60 for men. While these were raised in 6-month increments under 1996 legislation, the increase was unusual as workers could still retire at earlier ages with a reduction in benefits which were subsequently topped-up to full value upon reaching the age of normal retirement. The normal pension formula was extremely generous at 60 percent of the highest past wages averaged over 12 months for workers with full service (20 years for women and 25 years for men). Above that, the base was increased by one percent for each year of service above the minimum. In many occupations, more than one year was credited for each year of service
Pensions were paid to 2.8 million persons in mid-1996. Of these, some 2.1 million old age pensions were received. About 19 percent of the old-age pensioners received pensions provided on favorable terms with supplementary years of credited service. Further, these pensioners received higher than average pensions, as outlays represented 23 percent of all old-age payments. While theoretical replacement rates based on the law were very generous, because inflation indexing did not keep up with wage growth, actual average replacement rates were much lower than the formula would suggest. In July 1996 average pensions were 36 percent of average wages. In fact, throughout the 1980s and 1990s, the average pension replacement rate ranged from a low of 24 percent of the average wage in 1992 to a high of 42 percent in 1991. These low pension replacement rates resulted, in part, from the financial constraints of the system.
In mid-1996, employers were making contributions on behalf of about 5 million workers. This meant that 1.8 workers were paying for each pensioner, an extremely low ratio compared to the demographics of the population. In other words, the system dependency ratio, that is the ratio of pensioners to contributors, was 0.56 in 1995 while the old age dependency ratio, that is the ratio of persons age 60 and over to the working age population (age 20-59 years), was 0.18. By comparison, in the United States, these ratios were 0.31 and 0.30 respectively. Kazakhstan, a country with very favorable demographics, had a high system dependency burden and correspondingly high payroll taxes due to ineffective collection procedures and early retirement ages. Essentially, the growth of the informal sector and the development of wage arrears created a relatively small tax base relative to the working population. In fact, estimates suggested that less than half of all potential contributions were actually collected. This, of course, is not just a problem in Kazakhstan but affects virtually all CIS countries. In fact, tax collection shortfalls are much more severe in the CIS than in central and eastern Europe.
The system of collection and payment under the old pension system was inefficient and ineffective. The Ministry of Labor and Social Protection (MOLSP) and the local departments of social protection were responsible for the collection of pension fund revenues and the delivery of benefits. Each raion department of social protection had a pension fund department. That department was responsible for the auditing and enforcement of contribution collections. The monthly contributions were deposited into two accounts. Seventy percent of collections went to a raion account and the remaining 30 percent was supposed to be deposited into a central account in Almaty to be used for reallocation to the oblasts. The actual result of this arrangement was that funds remained at raion level until all local pensioners were paid. Further, relatively well-off oblasts, such as Almaty, appeared not to actively enforce collection compliance once they had sufficient funds to pay their own pensioners. While collections under this system were aided by local self-interest, the inefficiencies in allocation demanded administrative reform. Local authorities tended to use pension fund revenues to pay for family allowances under the premise that these funds would be reimbursed by local budgets. Since local funds were generally insufficient, these expenditure frequently were never repaid.
Under the former PAYG system, considerable arrears in collection were built-up on the part of contributing employers, with the duration of back-payments differing considerably by oblast. Prior to reform, the accumulation of arrears -- both in terms of payments and contributions -- was growing. On January 1, 1996 contribution arrears from enterprises totaled 40 billion Tenge, 26 billion Tenge higher than one year earlier. Contribution arrears from local and republic government ministries and organizations amounted to another 2.3 billion Tenge. By July 1, 1996, reported contribution arrears from enterprises were 49.6 billion Tenge -- equivalent to five months of pension payments at the time. Further, even if these arrears were made up, the pension fund only collected revenues from 5 million out of an estimated labor force of 7.8 million. Consequently, collections only amounted to 45-52 percent of potential revenues.
The build-up of contribution arrears and the general state of system non-compliance led to a significant build-up in arrears for the payment of pensioners. The pension payment arrears due pensioners were 26 billion Tenge at the beginning of the 1996 and peaked at 32 billion Tenge at the end of June 1996. As a result, government needed to transfer 36 billion Tenge from the state budget to the pension fund in 1997 to cover payments, including paying down the arrears, covering the deficit for the rest of the year, and paying for other administrative expenses. The build-up of back pensions began to be a significant focal point for social unrest, which, ultimately, opened a window of opportunity for the enactment of the current Kazakhstan pension reform.
III. An Outline of the Pension Reform
The new system has been a challenge to the administrative capacity and imagination of the Kazakhstan government as it creates an entirely new industry and entirely new regulatory responsibilities. Several new institutions play a key role in the functioning of the new system. These include the state accumulation fund (SAF), non-state accumulation funds (NSAFs), AMCs and custodian banks.
Each employee and self-employed worker can choose to become a member of either the SAF or a NSAF. Under pension reform, participants can contribute to private NSAFs or the SAF. The SAF was offered as an alternative to funds run by the private sector. In Kazakhstan, there is mistrust of both the state and private-sector financial organizations, particularly after problems arose with initial privatization funds. Further, examples of fund collapses, e.g. the 3M fund in Russia, are well-known in Kazakhstan. Thus, the SAF was started so employees wary of NSAFs could select a fund in which they had greater trust.
SAF management is the responsibility of the Government. Each NSAF must hire one AMC; each asset management company can manage assets for multiple funds. While the NSAF may direct the overall allocation of assets, the fund’s AMC is in charge of making day-to-day transaction decisions. Each fund keeps the accumulated assets of fund participants exclusively with one authorized bank custodian which accounts for and reports on all transactions, portfolio allocation and investment returns. The custodian bank also often acts as the broker for the transactions. The basic tri-partite structure was instituted to provide for a clear separation of accounts and responsibilities so that a system of checks and balances could thwart any fraud and abuse.
The allocation of investments by the SAF and the NSAFs are closely controlled. Both types of funds must invest a minimum of 50 percent of their assets in Government securities. The SAF, by law, can invest up to 40 percent of assets in designated national (state-owned) banks deposits and up to 10 percent in the issues of international institutions such as the IMF and IBRD. The NSAFs have greater flexibility in their investment portfolios as they may invest a maximum of up to 30 percent in Class A corporate securities. Class A securities are listed on the Kazakhstan Stock Exchange and have had at least one year of financial statements audited in accordance with international standards.
Three Government authorities are responsible for regulating the various elements of the pension system: (i) the NSC for the asset management companies; (ii) the NPA for the pension funds, reporting to the Ministry of Labor and Social Protection (MOLSP); and (iii) the NBK for the custodian banks. The State Pension Payments Center (SPPC), also reporting to the MOLSP, is the key administrative institution for the residual state system.
Although the new pension system officially started in January 1998, many steps are needed to complete the reform as entirely new policies, institutions, and regulatory capacities are being developed and implemented. The reform process will continue on an ongoing basis as the system develops and begins to function efficiently. Because the development of institutions is a process and not an event, the actual implementation and development of the pension reform will continue to be improved over the years. This has certainly been the case in Latin American countries which have instituted funded systems. The successful implementation of the reform will provide for the retirement income security of future generations of citizens and will promote the development of the financial market and the private sector in Kazakhstan.
IV. State Accumulation Pension Fund.
The Government of Kazakhstan is the founder of the SAF. It guarantees the safety of the contributions to the fund. It is established as a closed-end joint stock company with a capital of 10 million Tenge allocated to it by the Ministry of Finance. Participants may select the SAF as their pension fund of choice. Contributions also flow to the SAF from participants who have not specifically designated a NSAF. The MOLSP established the SAF, and together with the MOF, is jointly responsible for approving the supervising committee. The asset management company for the SAF is the National Bank of Kazakhstan (NBK), that also serves as the custodian bank.
The percentage share of SAF holdings has decreased throughout 1998. As of January 1, 1999, the NSC reported total assets in pension accumulation funds amounting to 23.6 billion Tenge. At that time, just over 76 percent of total assets were invested in the state pension fund. Some projections suggest the SAF will only hold around 70 percent of total pension assets by the end of 1999. This is consistent with the philosophy behind the reform which would have the share of the SAF steadily decline so that by the end of the decade it would no longer be predominant in the market. When that happens, the SAF may be privatized.
By December 1998, the SAF had transferred some 600 million Tenge to the NSAFs for individuals who chose to join the non-state funds. These monies had been automatically deposited with the SAF during the opening months of the reform when employee decisions had not all been formalized. By December 1998, some 300 million Tenge still were waiting to be transferred. The SAF has also made refunds on behalf of about 900 individuals who have either left the country or who have died since January 1998.
At this point, because the SAF is not required to use a ‘for profit’ AMC, it has a cost advantage with regard to the NSAFs. Further, the establishment of the SAF is subsidized in a number of ways by the Government (in terms of office space and salaries). In the future, the diversity of investments permitted the NSAFs ought to provide higher rates of return to those funds and, consequently, better investments for participants. In the near term, however, investment instruments have been limited as the stock market has developed more slowly than expected. Hence, the SAF clearly has a cost advantage which should be addressed if its share of the market is to decline. One way to equalize some of these differences would be to allow the NSAFs invest 5 to 10 percent of fund assets in international equities.
The SAF has now taken over the function of recording and maintaining its own data base of revenues and accumulations from the SPPC. This is a positive step and should give the SAF better control of its own financial position. In taking over the data base, the SAF identified numerous problems with the way in which the information was held by the SPPC. The SAF is currently working to improve the quality of the data and the accuracy of the information. The next step for the SAF would be the purchase of its own computers. The SAF also needs to employ directly staff capable of assessing its computer needs and able to instruct and control its suppliers of computer hardware or software.
V. Non-State Accumulation Funds
NSAFs are private-sector entities established as closed-end joint-stock companies. These may be either open-ended or corporate. A minimum charter capital requirement of 55 million Tenge is required for open funds and 10 million Tenge for closed funds. Minimum capital requirements will be increased to 90 million Tenge for open funds, and increased to Tenge 20 million for corporate funds as of April 1999. The authorized use of the charter capital, however, needs to be revisited with a more precise definition distinguishing charter capital from operating capital and reserves. Further, consideration should be given to instituting a capital requirement as a minimum percentage of fund assets to ensure that sufficient funds are available to supplement earnings below legal performance guidelines no matter the size of the fund.
Legal entities of the Republic of Kazakhstan can be the founders and shareholders of the corporate pension funds and of the open pension funds, along with individuals. Citizens can be fund managers upon passing a qualification test. Pension funds have the right to receive commissions and are obliged to contract with one pension management company. The contract defines the distribution of commissions between the NSAF and the asset management company (AMC) for the management of fund assets, the method of transferring funds for benefit payments, and includes the rules of the contract on trust property management.
Twelve non-state accumulation funds were licensed as of February 1999:
Umit Pension Fund, (open)
Valyut-Tranzit Pension Fund (open)
D.Kunaev Pension Fund (open)
Narodny Pension Fund (open)
Kazakhstan Pension Fund (open)
Narodny Bank Pension Fund (open)
Kazakhmys Pension Fund (corporate)
Ular Pension Fund (open)
KaspiiMunaigas Pension Fund (open)
Nefte Gas-DEM (open)
Kazakhstan Trade Unions Federation (open)
In November 1998, among the NSAFs, the Narodny Bank had the largest share, at just over 50 percent of assets. There was further industry concentration with the Narodny Bank fund and the next six largest of the 11 (at that time) private funds accounting for 94 percent of total non-state fund assets. A number of observers have suggested that ultimately only five or six non-state funds will remain in the market. Given the significant economies of scale that are observed in pension plan operations, these forecasts are likely to be realized. (For example, the Zhardem pension fund was reorganized and merged with D.Kunaev’s pension fund. When Zhardem’s temporary license has expired, no new license was issued to it by NPA).
The fee structure for the pension funds is set at up to 1 percent of contributions and up to 10 percent of the investment income. The architects of the system hoped to avoid some of the excessive fees generated in other privatized systems resulting from churning of accounts from fund to fund due to aggressive marketing. Such churning had characterized the funded system in Chile at the expense of the return to affiliates. For this reason, the drafters of the Kazakhstan legislation decided to institute a maximum fee structure. These fees may be divided between the NSAFs and the AMC. The AMC share is set at 0.15 percent of contributions and 5 percent of investment income.
Initially, there was considerable discussion about the adequacy of the fee structure in terms of fund profitability. Recently, renewed discussion is taking place with funds indicating that they need 100,000 to 150,000 participants to break even. The total number of contributors in the private funds is 600,000 with no fund except the Narodny Bank at the 100,000 level. Of course, it was anticipated that for the first few years pension funds could expect to be operating at a loss until they built up their participant and asset base. And, if the share of the SAF falls, a number of funds should be able to include 100,000 participant. Another recent development that could improve the contributor base for the smaller pension funds is the proposed use of the SPPC network on a commercial basis for record-keeping, contributor notification and other services. In that way, the private funds could become more competitive with the SAF and the Narodny Bank which started out with established networks. Both the issue of fee structure and SPPC commercial ventures require additional discussion.
The ownership of the fund assets remains a continuing topic of discussion. As there is no trust law in Kazakhstan, personal pension accounts are not clearly the property of the participant. This ambiguity could place individual accounts at risk should a fund be in financial difficulties on its own account and look to the assets of the participants as a buffer. This is an issue that has been discussed extensively and appropriate solution is required in the very near future before a problem arises that could affect the confidence of the population in the pension system.
VI. Asset Management Companies (AMCs)
The AMCs are legal entities created in the form of closed-end joint stock companies. The minimum charter capital required is 80 million Tenge. Any citizen can be a manager of a company upon certification. The AMC has the right to sign contracts with one or more NSAFs. The company’s equity must, however, increase in proportion to the assets under management. Four private asset management companies have been licensed as of June 1998 but one has had its license suspended by the NPA following an inspection. The assets of that AMC were transferred to other asset managers. As of February 1, 1999, six companies (not including the NBK for the SAF) have active licenses for pension asset management:
ABN-Amro Asset Management
Zhetisu Asset Management
Narodny Bank Asset Management
Aktiv-Invest Asset Management
Ak Niet Asset Management
The AMCs have provided the NSAFs with extremely high rates of return from their investments as National Bank notes and T-Bills have been offering interest rates of 23-25 percent. As a result, even after fees and commissions, pension fund participants can expect to obtain total returns of well over 15 percent for 1998 on the whole. In addition, towards the end of 1998, AMCs started purchasing back Kazakhstan Eurobond issues to diversify their portfolios. Unfortunately, the combination of the Asia crisis, the Russian meltdown, and the substantial fall in commodity prices has had an adverse impact on the development of the capital market, and, as a consequence, the possibility for the NSAFs to acquire more diversified portfolios.
For example, NSAFs can invest in shares of Class A companies, those which have been subject to an independent audit based on international standards. Shares of Kazakhstan companies were to be made available through the Blue Chip privatization in which majority blocks of shares were first to be tendered to strategic investors and then gradually sold through the stock exchange. By these measures, the government was to achieve its multiple objectives of privatization, securities market development and pension reform by selling its remaining shares in large enterprises on the Kazakhstan Stock Exchange (KSE). This approach, implemented gradually, taking into account the budgetary needs for privatization revenues along with the pension funds’ ability to absorb equity shares, was to provide the pension funds with the opportunity to acquire quality Kazakh companies in a competitive manner.
Unfortunately, the Blue Chip program was first delayed by the authorities and then the Asian Crisis created problems with the ability of some of the strategic investors to meet their obligations. With the Russian crisis and the precipitous declines in all emerging market stock exchanges, the Blue Chip program was further delayed. Currently, the sale of 15-20% of shares in four large enterprises (Aktobemunaigas, Ust Kamenogorsk Titanium and Magnesium, Mangystaumunaigas, Kazakhmys) is scheduled through the KSE, and a program developed to sell the remaining shares in these and other large enterprises over the coming years. A tender has already been announced for additional four “blue chip” companies (Kazakhstan Aluminum, Kazchrom, SSGPO, Kazzinc), and a winner is scheduled to be selected in early March 1999. According to the schedule, the shares of all eight companies shall be issued in the 3rd-4th Quarter 1999. These shares would be purchased by the AMCs for the accounts of the NSAFs. In recent months, there has already been some small diversification in investments with a small amount of equity and corporate bond investments added to NSAF portfolios.
VII. The Custodian Banks
The custodian banks are the lynchpin of the pension reform system ensuring that neither the NASFs or the AMCs can make felonious use of the assets in the accounts of plan participants. The National Bank of Kazakhstan has licensed seven custodian banks (not including the NBK for the SAF). They are the following:
Almaty Merchant Bank
These banks have been licensed with care ensuring that they are credible depository institutions. Nonetheless, the NBK will need to monitor their activities, particularly if it appears that there is the possibility of interlocking financial interests between the NSAFs, the AMCs and the custodian banks. At this point, there are close ties between a number of NSAFs and AMCs through interlocking founders. This is not necessarily a cause for concern, although initially it was presumed that the NSAFs, the AMCs, and the custodian banks would all be independent of each other from a business perspective. Prior to the passage of the Pension Reform Legislation, there had been discussion as to whether there should be separate NSAFs and AMCs or whether the AMC function should be combined with the pension fund as is done in Chile. Should the custodian bank also be closely linked to the other two entities, however, the potential for fraud would increase, if only through the potential for party-in-interest investments which would benefit the founders in question but would not maximize returns for the plan participants.
VIII. The Regulatory Structure
National Pension Agency (NPA).
The NPA is responsible for: (i) licensing and supervising NSAFs; (ii) approving pension contract forms; (iii) regulating benefit payments, confidentiality of individual pension accounts, inter-fund account transfers; and (iv) regulating the operation, accounting, reorganization, merger and liquidation of funds. It is also responsible for control and supervision of SAF activities.
Under its present leadership, the NPA is increasingly upgrading its performance as a pension fund regulatory agency, issuing and revising important regulations for the operation of pension funds and planning ahead. The NPA has completed inspecting all 13 licensed pension funds, including the SAF. As a result, after the first year of operation, 8 non-state accumulation funds have already received general licenses for an unlimited time, instead of initial temporary ones. As expected at the beginning of any pension reform process, many violations were found, some of them serious. NPA has indicated corrective measures and will consider the application of sanctions as appropriate. New regulations have been issued on accounting and reporting procedures, independent external auditing, the reorganization-liquidation of pension funds and minimum capital requirements.
There are a number of issues that the NPA will still need to address. For example, reporting requirements should be simplified. NSAFs currently report that they are a burden, and the limited staff of the NPA may not have the capacity to process them, in any case. In addition, regulations on mergers need to be revised to clarify the transfer of pension assets from one pension fund to another. The NPA should also start to consider rating pension funds, coordinating with the NSC for the simultaneous rating of AMCs.
National Securities Commission
As regards its role in new pension system development, the NSC is responsible for: (i) licensing and supervision of AMCs; (ii) licensing and supervising custodian banks (with the NBK); and (iii) regulating the investment activity of NSAFs. The NSC has established a separate pension unit for this purpose. The NSC is increasing its efficiency in its role as regulator; the license of one AMC was suspended as a result of an inspection showing unsatisfactory accounting practices.
There will be many more issues to address, however, as the pension system matures and consolidates. First, the NSC reporting requirements are also considered excessive and the capacity of the NSC to make efficient use of the information is an open question. The NSC should coordinate with the NPA on instructions in case of pension fund mergers and transfers of pension assets. For example, the current practice of requiring AMCs to realize the pension assets with a discount is not in the interest of the contributors. Pension assets should be transferred as invested. The establishment of a permanent working committee of the three regulatory agencies would be one way to coordinate reporting requirements using common standards and sharing information.
The NSC should also institute on strategic planning for future permissible investments in dialogue with the AMCs. For instance, the authorized investment in Eurobonds requires instructions on the calculation and accounting of these investments. Investments in corporate bonds require complementary regulations on the type of bond investments allowed, the calculation of returns, and accounting and reporting procedures.
National Bank of Kazakhstan
The NBK is responsible for licensing and regulating the custodian banks which safeguard the assets of the NSAFs. The NBK has assigned the Banking Supervision Department critical responsibilities for pension oversight. The Department has gained a fine reputation for managing and modernizing the current banking system. To date, these functions appear to have been managed smoothly, and the NBK appears to be well received. Additionally, the NBK has a separate unit acting as asset manager for the SAF. The NBK is also the custodian bank for the SAF. In theory, these functions are to be phased out and commercial asset managers and custodians substituted. The way this process actually unfolds will be crucial to the eventual complete privatization of the pension system.
IX. The State Pension Payment Center
The basic role of the SPPC should be threefold: (i) to calculate and pay PAYGO pensions; (ii) to ensure that the working population participates in the system; and (iii) to ensure that contributions are accurately assigned to the fund chosen by the contributor. In addition, to support the new pension reform legislation, the responsibility for the payment of current pensions has been transferred to the State Pension Payment Center (SPPC).
The role of the SPPC has been changing since its inception. Initially, the SPPC was responsible for maintaining the SAF data base. This role has now been taken over by the SAF. The SPPC is now contemplating offering pension funds access to its network of offices on a commercial basis to register workers and provide information to participants. The accounting and control systems for this proposed mixture of mandatory and commercial services will need to be separated to reflect these differences.
Further, both the payroll taxes and the funded contributions used to be forwarded to the SPPC. The SPPC subsequently transferred the 10-percent contributions to designated pension funds. Payroll tax enforcement responsibilities were given to the state revenue collection service. These procedures were changed in 1999, however, with the institution of the new general revenue (not earmarked) payroll tax. As a result, the SPPC is no longer responsible for the collection of payroll taxes at all, as they will be deposited directly to a state account. The SPPC will continue to receive the 10-percent contributions from employers and will direct them to an appropriate fund for each employee.
The payment of pensions will continue to be one of the basic functions of the SPPC. Pensions have been paid each month within one month of the due date, and, as of the end of November 1998, all pensions for October 1998 had been paid. Timely payment of pensions is a condition for the release of each tranche of the World Bank’s Pension Reform Adjustment Loan (PRAL). Gradually, the SPPC is taking on responsibility for the calculation of pensions as well. As of January 1, 1999, it took on the added responsibility for the calculation of military and state-body pensions. It is anticipated that it will take over the calculation of all pensions within the next few years.
The SPPC is also responsible for issuing Social Individual Codes (SICs) to ensure that all persons of working age have an SIC with which to register their contributions to the funded system. Because this operation was initially hampered by delays in the installation of the central and oblast/raion computer system, the World Bank initially reduced its conditionality for the Loan for SIC registration to a very low 10 percent of persons identified as working in the formal sector. The trust of the Bank was not misplaced, however, as now approximately 3.2 million SICs have been issued to workers in the formal sector.
Problems remain as a number of duplicate SICs have been issued and more work is required to issue SICs to persons outside the formal sector. Discussions have been held to amalgamate the SIC with the tax code, although no agreement has been reached. Such coordination would greatly improve the tax authorities’ ability to enforce tax collection. For example, the personal income tax base of the tax authorities ought to be reconciled with that of the SPPC through the institution of one identification number. While Government undoubtedly underestimated the importance of the SIC at the start of reform, there is evidence that through a learning process, the necessity to issue SICs to everyone in the population is now taken extremely seriously. There are plans for Government to use information from the 1999 Census of the Population to help in the registration of the working-age population.
One of the major drawbacks of the former PAYG system was its inability to collect much more than half of all collections due. To change this situation, modern tax collection techniques are to be adopted by the tax inspectorate, including the linking of information available to the SPPC and other agencies to identify individuals and businesses who avoid or evade their payment obligations. Similarly, the linkage of contributions to benefits should provide incentives for individuals to contribute to the funded plans. While the tax authority now has responsibility for collection, the SPPC still has a significant role in the identification of potential taxpayers. To date, the SPPC has been a passive partner. This must change in the future so that the tax inspectorate has all information available to target compliance and conduct random audits.
The PAYG pension system was based on soviet accounting practices which are not in keeping with international standards. These practices, which are not transparent, in combination with the decentralized collection and payment of benefits and minimal reporting procedures, prevented any reasonable degree of fiscal control by the MOLSP. To ensure future transparency and fiscal accountability, an independent assessment is needed to introduce modern accounting and auditing standards to the SPPC. Following their introduction, there is to be an independent audit of SPPC finances.
The efficient transfer of contributions by employers on behalf of employees and the self-employed is necessary to build public confidence in the system and to maximize the returns participants receive on their contributions. The SPPC has improved the time taken to record and transmit contributions to the accumulation funds. In cases in which there are no substantial errors, the SPPC transmits all payments received before mid-afternoon on the day of receipt and other payments within one day. These improvements have resulted in significantly less money held in transit within the SPPC. However, because the SPPC achieves this by returning all errors to the enterprise, this action still delays transmission to the accumulation fund. The SPPC needs to take a more active roll in resolving mistakes to speed up transfers.
X. The Benefit Structure
Initial estimates by Government suggested that a 10-percent contribution would provide a replacement rate of around 60 percent for career workers. For the first few years, the distributions from the new pension system will be quite small, however, and will be provided as lump-sum benefits. Regulations on the payment of pensions from individual accounts are to be promulgated in the future, in conjunction with the development of the insurance industry. As Government expects the pension system to provide adequate future retirement income relative to earnings, ongoing analysis of the level of pensions by age cohort, gender, and length of working life is to be instituted. Such analysis is essential as pensions that rely on personal saving will vary across cohorts depending on system returns and may lead to lower replacement rates for women or other groups in the population.
Whether or not replacement rates meet the 60 percent replacement rate target depends on the period over which the contributions are made, the rate of return to the participants’ pension accounts, and the length of the contributory period. A simulation model using slightly different assumptions than those used by Government indicates that under reasonable assumptions, individual replacement rates may be less than the target 60 percent rate.
For example, Table 1 present estimates of total pensions (state plus funded) as a function of age cohort and work history. In the model, the individual contributes 10 percent of his or her wage from the start of the reform until retirement and receives a price-indexed pension from retirement until death. Administrative costs are assumed to take up 10 percent of contributions, savings earn a 3.5 percent rate of return, real wages grow at 2 percent while inflation is zero. Individuals start work at age 20 and retire at the official retirement age.
Table 1: Initial Replacement Rate (%)
as a Function of Work History and Starting Age
Work history: Age at start of reform:
25 years 35 years 45 years
40 years 30 (2.50) 34 (2.82) 44 (3.65)
35 years 27 (2.24) 31 (2.58) 36 (3.00)
30 years 24 (2.02) 28 (2.35) 33 (2.76)
20 years 18 (1.48) 22 (1.80) 30 (2.50)
Note: The figure in brackets is the initial pension as a function of the minimum pension.
Two main points emerge from these simulations. First, under the specified assumptions, initial pensions will typically replace less than 60 percent of the final wage. If workers want to have a higher replacement rate, they will have to save more voluntarily. Second, those with short contribution histories (which are likely to include women) may not be able to accumulate enough in their pension accounts to support themselves adequately in their old age. The reform could put these groups at risk of old-age poverty.
While current rates of return to pensions are currently far higher than initial Government assumptions and much higher than the return assumed in the simulation model, these high rates are not likely to last in perpetuity. And, as the stock market emerges, funds can expect fluctuations in returns. Fluctuations in emerging markets are still likely to be greater than those in developed market economies. Two main policy implications emerge. First, it will be important not to raise false expectations about the level of pensions the reformed system will support. Second, it may be necessary to raise contribution levels and make other changes in order to achieve social protection outcomes to approach the 60 percent target.
As part of pension reform, pensioners with full years of service (equal to 25 years for men and 20 years for women) will be eligible to receive a minimum pension if their combined plan pensions fall below a minimum consumption standard. A minimum of 2,400 Tenge per month was enacted under the 1998 budget law. That level is roughly equivalent to 70 percent of the 1997 survival minimum cited in the World Bank’s poverty assessment, Kazakhstan: Living Standards During the Transition. Allowing for the fact that pensioners may have other sources of income, including self-production of food and assistance of relatives, a target of 70 percent of the survival minimum appears both adequate and reasonable. Such benefits should be indexed for inflation automatically to ensure pensioners that with a full working history are not at risk of poverty.
Disability and Survivors’ Benefits.
Under the new legislation for social allowances, disability and survivors’ benefits have ceased to be insurance benefits and have become flat-rate allowances unrelated to service- or wage-history. These allowances are funded directly through the budget and are no longer a part of the pension system. Workers eligible for invalidity benefits, whose conditions preclude any possibility of work, face a sharp reduction in their income if they earn above the average wage. Similarly, low-wage earners in other categories may receive higher income from disability allowances than they did prior to disablement. For these reasons, these policies represent interim measures instituted to break with the former system. The longer-run goal for disability and survivor’s benefits is the provision of annuities through private disability and life insurance.
The emergence of a strong insurance industry is thus necessary to support the funded pension system. As the system matures, the payment options approved by government for the funded system (such as a fixed schedule of payments, single life annuities, joint and survivor annuities) will be in the hands of commercial insurers. Similarly, an extension of reform to mandate insurance-based disability and survivors’ benefits must be predicated on the emergence of a well-regulated insurance industry.
XI. The Costs of Pension Reform
The fiscal sustainability of the pension reform has been and is a crucial consideration. The fiscal framework developed by the authorities for 1998 after the passage of the pension reform legislation was to accommodate the increased budgetary expenditures incurred under pension reform through an overall deficit for 1998 of 5.5 percent. (The overall fiscal deficit, excluding privatization receipts, was estimated at 7.8 percent of GDP for 1998.) The underlying adjustment in other recurrent revenues and expenditure items equalled a decline of 0.7 percent of GDP.) In terms of the cost of transition, projections based on Government’s actuarial model indicate that the implicit government debt would amount to 110 percent of 1997 GDP, which compares favorably to other countries in the world.
When the reform was enacted, actuarial and economic projections indicated that the shift to a fully funded pension system would result in an additional fiscal cost of pension reform of 1.7 percent of GDP in 1998. The cost of moving to the new system using the results of the actuarial model are presented in the Table 2. In principle, Government could borrow the full amount to finance the increase in the fiscal deficit from the domestic market without causing undue pressure on aggregate demand as losses in public revenue would be compensated by an increase in non-public sector savings. Further, the revenues from privatization were also intended to buffer the costs of the reform from the impact of the increased deficit stemming from the reduction in the payroll tax of the 10-percent that was transferred to the pension funds.
Table 2: Cost of Pension Reform
According to Government’s Actuarial Projection Model
(billions of 1997 Tenge)
State Budget 1997 1998 1998 Cost of
plus Pension Fund Former No Change New Pension
System in Policies System Reform
Retirement Pensions 89.4 92.4 91.4 -1.0
Social pensions / allowances 19.8 21.0 21.0 0.0
Military pensions 5.1 5.2 5.2 0.0
Gov't contrib. to accumulation funds 0.0 0.0 10.9 10.9
Total outlays 114.3 118.6 128.5 9.9
Net payroll taxes 48.4 48.7 28.2 -20.5
Gross inflows to SPPC 70.2 70.5 41.0 -29.5
less, gov't contrib. - own employees 21.8 21.8 12.8 -9.0
Net Cost to Government 65.9 69.9 100.3 30.4
Percent of GDP 3.9% 4.0% 5.7% 1.7%
The actuarial model shows that in the absence of reform, the costs of the existing system would increase slightly in 1998, from 3.9 to 4.0 percent of GDP, mainly as a result of additional pensioners. With reform, the cost of the new system would be 5.7 percent of GDP, or an increase of 1.7 percent of GDP. The cost of reform is mainly due to the reduction in the payroll tax rate from 25.5 percent to 15 percent in 1998, which was estimated to reduce gross inflows to the PAYG system by 30 billion Tenge. For the Government, as for all other employers, the reduction in payroll tax should be offset by the 10-percent cash contribution employers will be required to make on behalf of their employees to the private pension funds.
While the net cost of pension reform was estimated at 1.7 percent of GDP in 1998, it was projected to rise to 2.4 percent of GDP in 2008. The steady rise in the cost of reform was due to Government’s assumptions of further reduction in the payroll tax rate by one percent per annum to 5 percent in 2008. The incremental cost then began to decline to zero. In 2026, the gross cost of the new system was projected to be equivalent to 2 percent of GDP, considerably less than continuing with the former system. With the change in the financing of the system to general revenue financing, the projection of the costs of reform must be redone. The revenues stemming from the payroll tax will no longer be related to the reform. Nonetheless, it is possible to calculate the funds required if the system had not been reformed and the funds required to fund both the funded system and the residual system. This difference represents the costs of the reform. It is also important to keep in mind that at the time of the reform, the 25.5 percent payroll tax did not cover the costs of the old system. Government is committed to estimating the annual costs of the public system as part of the fiscal planning process.
Further, Government is committed to a program of fiscal adjustment, as is necessary, on a permanent basis. Certainly, other factors could influence adversely fiscal expenditures and macroeconomic balance. If Kazakhstan were to suffer an economic collapse like Russian, both the pension reform and the entire economy would be in a state of turmoil. And the impact of the Asian crisis, Russia’s financial crisis, and the collapse of commodity market prices, especially oil prices, have had an adverse exogenous impact on Kazakhstan’s economy. While, it is unlikely that the pension reform could have, or would have been enacted initially under similar economic conditions, those were not the circumstances in 1997. At that time the economy looked strong and world conditions favorable. Hindsight notwithstanding, to date Government has been taking prudent steps in both its fiscal and monetary policy to deal with projected reductions in revenues and in economic growth.
The pension reform has already developed a new financial sector in Kazakhstan through the blossoming of private pension funds and asset management companies. While the reform appeared to get off to a slow start, the expertise being developed is already substantial and significant assets have started to flow to the private system. These assets will build the foundation for a transparent investment process that will provide a firm foundation for business practices that are not tainted with the type of corruption that is rampant in many transition economies with the emergence of strong mafias and others engaged questionable business practices. As a result of the pension reform, the regulation of financial institutions is being strengthened and that will reduce the scope for fraud and abuse. The NBK is known to have developed into one of the most credible institutions in Kazakhstan. Many of the new regulators of the pension system have benefited from initial NBK training.
Further, the reform sends a strong signal to the population that times have changed and that individual responsibility must be counted on to replace the state in the provision of social needs. Such a change in mentality is key if the transition to a market economy is to succeed. While the progress of the pension reform in terms of portfolio diversification is likely to be slower given current economic circumstances, the reform should not be rolled-back or rescinded on the basis of budgetary concerns as this would surely damage the potential for economic growth.
XII. World Bank Involvement
Pension reform in Kazakhstan was designed and developed by Kazakhstan’s officials and legislators. As such, it is indigenous to Kazakhstan. While Kazakhstan officials were cognizant of the World Bank’s volume, Adverting the Old Age Crisis, they were not guided by World Bank technical assistance. Similarly, while the U.S. Agency for International Development encouraged a radical reform, the champions of the reform accepted that assistance only when they considered it to be productive. The new law was very much the result of a group of reformers who sought to implement a three-pronged development strategy announced in 1997 including privatization, capital market development and pension reform. Further those reformers understood that an incremental reform of the PAYGO system was not possible as it faced too much political opposition. In addition, as the reformers came from the financial sectors, the financial imperatives for the reform were stressed.
In March 1997, it became clear to the Bank that pension reform was underway. The Bank responded to the request of Government for comments on the initial design of the system. Our comments reflected a number of concerns including the benefit structure, the rapid pace of the reform, and the lack of financial sector and regulatory experience. Nonetheless, from the start our response was to support the basic structure of the reform as it was close to that suggested in Adverting the Old Age Crisis. Following the passage of the pension reform legislation in June 1997, a preparation mission was sent to Kazakhstan to assess the advisability of providing an adjustment loan and present recommendations to Bank management.
The Bank had one of two options. It could support the Kazakhstan pension reform with an adjustment loan, including stringent conditionalities to ensure that the reform was successful, or not support the reform on the premise that it could not succeed. A decision was taken to support the reform as the basic structure was felt to be sound if Government would agree to strong conditionalities and accept technical assistance provided though the Bank ensuring that those conditions could be met. Further, to ensure transparency, the mission team endeavored to document all the possible pitfalls of the operation and all the deficiencies of the reform to assist Government and to make all the risks known to Bank management. In other words, assistance was provided to the Government of Kazakhstan knowing that the pension reform was risky but deciding that it would be proper to support the pension reform policy as it would be an example to other CIS countries in either its failure or success. The end result of these deliberations was to grant the Republic of Kazakhstan the Pension Reform Adjustment Loan (PRAL) in 1998 combined with a Pension Reform Implementation Project (PRIL) funded as a component the restructured Finance and Enterprise Development Project (FEDP).
Since the enactment of the pension reform, considerable progress has been made in its implementation. Initially, that implementation appeared to be bound for disaster as regulatory offices were unqualified to act as regulators and computer equipment was insufficient to handle the barrage of data trying to make its way through the system. With time, Government officials began to realize the enormity of the task that had undertaken. With time, professionalism in both regulation and administration started to take hold. At this time, with the assistance of international donors, including the ADB, the World Bank, and USAID, the Kazakhstan pension system has the potential to be as successful as that of Chile. This is not to say that there are no further problems. The challenges are still tremendous, as are the challenges in many other areas of the Kazakhstan economy before it can enjoy full employment and sustained economic growth. Nonetheless, given the difficult economic situation Kazakhstan is facing, with external shocks compounding the strain of transition, the progress of the pension reform has been steady and further success can be expected.
Anderson, J.M., Kolberg, J.E., Vroman, W. and Evans-Young, T. 1997. “The Pension system of the Republic of Kazakhstan: Policy, Structure, Operations, and Reform.” Final Report to the Ministry of Labor and Social Protection, Republic of Kazakhstan. (April).
World Bank. 1998. Kazakhstan: Living Standards During the Transition. Report N. 17520-KZ (March 23).
World Bank. 1998. Report and Recommendation of the President of the International Bank for Reconstruction and Development to the Executive Directors on a Proposed Pension Reform Adjustment Loan in an Amount Equivalent to US$300 Million to the Republic of Kazakhstan. Report No: P 7242 KZ (July 3).
World Bank. 1994. Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth. Oxford University Press.
Paper Prepared for the World Bank Pension Primer
On Country Experiences in Europe and Central Asia
March 17, 1999