Indexation of Old Age Pensions Jaroslav Vostatek Abstract Indexation of Czech public pensions is not comprehensible already because the basic and percentage pension assessments are indexed separately and the indexation rate of the percentage assessment depends on indexation of the basic assessment. This incomprehensibility is a cause of attempts to make the indexation flat, which is in contradiction to the principle of equivalence in insurance-based systems. Such tendencies show evident efforts to correct the level of solidarity contained in new pension entitlements via pension indexation. The goal of the paper is to propose a variant comprehensible system of indexation of Czech old age pensions, preferably related to possible alternatives of pension reform, both total and partial, related only or mainly to the indexation of pensions. In that process, we will build upon foreign knowledge and research work regarding pensions and related benefits, especially the means-tested housing benefit. We will also deal with potential feedback of pension indexation for the future shape of the existing Czech system of public pensions. Keywords: flat-rate pension, housing benefit, indexation, old-age pensions, solidarity JEL Classification: H55, H53, I38 1 Introduction Sobotka’s government adopted two amendments to the legal regulation of pension indexation. Since 2015, it has re-established the scheme of public pension indexation according to which these pensions fully reflect the index of consumer price growth and from one third the index of real wage growth. However, it was not a simple return to the previous regulation, because by the “small” pension reform Nečas’ government changed also other parameters of pensions having an immediate impact on pension indexation. The whole construction of indexation is thus considerably more complex than before when, by means of a decree, governments were allowed to change the relation between the basic and percentage assessments of pensions at annual indexation. The small pension reform introduced parametrisation of the basic pension assessment at the level of 9% of the national average earnings (NAE), or more precisely the general assessment basis for calculating new old-age pensions, so that since that time the only task of the government at annual increase of pensions is to approve correctness of calculation of 9% of NAE. Up to this point, pension indexation is very simple and automatic – but only as to the basic pension assessment. Automatic indexation of the basic pension assessment depends on development of the nominal wage in the economy – in the highest possible extent (100%). The basic pension assessment is usually a relatively small part of the total old-age pension. In relation to the average old-age pension it represents only 21–22%. A substantial part of the total old-age pension is its percentage assessment. The present calculation of indexation of the percentage pension assessment is overcomplicated, because the index increasing the percentage assessment is determined in such a way that in case of an average old-age pension a total of the increase in the basic pension assessment and the increase in the percentage pension assessment corresponds to the increase of an average old-age pension defined at the level of a total of price growth and one third (a half from 2018) of the average real wage growth. We can only guess a reason for such a complicated method of indexation of the percentage pension assessment: a ceiling put on a pension expenditure increase. However, with regard to a relatively low share of the basic pension assessment the fiscal risk is not excessively high. Moreover, the law amendment prescribed for 2015 that the increase must be at least 1.8% for an average old-age pension. It can thus be concluded that there is no factual reason for such a complicated method of indexation of the percentage pension assessment. In June 2017, the President of the Czech Republic signed an amendment law on pension insurance which increases the inclusion of nominal wage growth in percentage assessment indexation from one third to a half and, in addition to that, it improves calculation of the inflation index: besides the existing consumer price growth index, it will also calculate the index of consumer prices (costs of living) for households of old-age pensioners and will use the higher of those price growths in indexation. This pension insurance law amendment also contains a ceiling on the increasing statutory retirement age at 65 years and a regular assessment of need to increase this age. In December 2016, the Premier Sobotka charged the Expert Commission for Pension Reform (Odborná komise pro důchodovou reformu, OKDR) to seek possible ways of changing the indexation scheme which would lead to a reduction of big differences of final indexation amounts depending on the current pension level. In May 2017, an OKDR work team described the problem as a high differentiation of indexation amounts at application of the then and newly approved mechanism of pension indexation which in the long run may result in a hardly respectable low indexation of low pensions and scissors opening between low and high pensions. The flat indexation at which all pensions increase by the same amount was rejected to become the solution to the problem. Similarly, an additional introduction of an absolute minimum of the increase in the percentage pension assessment at the expense of the percentage assessment of the other pensioners was rejected, as it would probably result in “adverse reactions by the general political spectrum and public” and in technical problems due to its complexity which may “pose a threat to a timely determination of the level and payment of the indexations” (OKDR, 2017). The team advised the Premier to introduce a partial flat indexation of the percentage pension assessment, at the expense of the total increase of the percentage pension assessment, or to additionally introduce the absolute minimum of the increase in the percentage pension assessment which would be financed from extra funds. The OKDR’s proposals for adjustment of pension indexation were not based on an analysis of the current entire system of Czech public pensions in the context of the entire social policy, let alone international comparisons. Even though it is evident that pension indexation policy often pursues narrow and time-limited political goals in foreign countries as well, it is possible to see tendencies here that fall into corresponding pension systems. We will deal with this topic in the next part of the paper – to lay a foundation for our analysis aiming to find alternative systems of indexation of our old-age pensions. 2 Czech Public Old Age Pensions Czechia has undergone a major pension reform from 1996 and since this reform the public pension pillar has been called “pension insurance” probably because at that time we had had the pension insurance premiums already for three years. The tax reform (1993) introduced the pension insurance premiums as part of social security premiums; the tax reform authors clearly expected introduction of a comprehensive social insurance system of a Bismarck, Christiandemocratic type, with earnings-related benefits, separated from the state budget. A subsequent social pension insurance scheme has never been implemented in our country; the pensions have been financed from the state budget – since the early 1950s – and the “re-established” pension insurance premiums are in fact a pension tax which is largely redundant – from the systemic point of view. Even the state budget has no “pension account” which would anyhow regulate pension expenditures of the state budget depending on development of collected pension insurance premiums (MF, 2012). Pensions provided in accordance with the law on the pension insurance are state pensions and the scheme is the (largest) public expenditure program – with all the possible advantages and disadvantages of this structure. The practical result of this situation is that although state budget expenditures for state pensions are mandatory expenditures of the state, they can be adjusted even every year, e.g. pursuant to the state budget law. In case of a threat of an excessive deficit of the state budget, pension expenditures can be – depending on the government policy – reduced. Should such a situation happen, what is most likely to “be on the sharp end” is the (usually annual) pension indexation. This often happens in countries with state pensions; we also experienced it in our country in previous years. Although a given country has rules for automatic indexation in place, e.g. depending on price development, they can be suspended in the respective fiscal year. However, everything can be even the other way around. An alternative to the system of tax-financed state pensions consists in systems of social pension insurance the funding of which follows separate, autonomous rules, relatively independent of the government and its fiscal policy in the respective year. In many countries with a social insurance scheme, these systems are co-financed from the state budget and to a certain degree analogous situation can occur there as in the countries with state pensions. In some countries, however, the system of social old-age pension insurance is in fact thoroughly separated from the state budget and in case of emergency it is precisely determined (by law) how a sudden lack of receipts from pension insurance contributions should reflect in the adjustment of pension entitlements (e.g. on clients’ accounts) and in indexation of pensions paid. The Swedish “automatic balancing mechanism” of this type also uses a basically “inherited” reserve fund, which is also a benefit from the perspective of investment revenues as a source of the social insurance fund; however, it can also be (and it was!) a serious disadvantage at the time of a financial crisis, or when the value of assets of the reserve fund deceases: several times in the past this led to a negative indexation of the Swedish “income pensions“. Since 1996, all the Czech pensions consist of two parts: a uniform “basic” pension assessment and a “percentage” pension assessment partially dependent on previous earnings. The basic assessment is a result of transformation of the previous state compensatory contribution, introduced in 1990 as a partial compensation of consumer price growth because of cancelation of retail price subsidies. The percentage assessment of the old-age pension adopted the previous technique of cutting higher earnings at the calculation of the individual assessment basis; however, it employed it considerably more towards the pension equalisation, in comparison with 1989 and previous years. The system of reduction thresholds (bend points) and related reduction coefficients, valid for calculation of the percentage assessment of old-age pension in our country since 1996, shows a significant similarity with the U.S. public pension system. The U.S. public pension system is of Beveridgean type, with prevailing quality of the solidary pension pillar. The Czech old-age “pension insurance” scheme is of the same type, with solidarity prevailing over equivalence roughly in a ratio of 2:1. However, the U.S. system is substantially simpler and – transparent; U.S. pensions are indexed every year depending on a special price index of pensioner consumer basket and I have not noticed any politicians’ discussion about indexation. In 2010, the Constitutional Court of the Czech Republic issued a ruling which states literally: The complicated construct of the pension system is so non-transparent that it is in fact incomprehensible to beneficiaries. (Constitutional Court, 2010). The Constitutional Court evaluated the then reduction thresholds and reduction coefficients as anticonstitutional, since in consequence of it and in combination with the other parameters and the existing structure of the pension system, they fail to provide a sufficient guarantee of the constitutionally guaranteed right of the corresponding material welfare and result in an unacceptable inequality among various groups of old-age insureds. For that reason, the section 15 of the law on pension insurance scheme, containing the reduction thresholds and coefficients, was repealed by court, as of 30 September 2011. Evaluating it from the economic point of view, it is evident that the Constitutional Court used the name and the terminology of the pension insurance law as a basis. If something is called insurance, then it should be insurance, in the respective case social pension insurance. It means that a pension should – fundamentally – depend on insurance contributions paid or on earnings based on which the insurance contribution and the pension are calculated. From the perspective of the pension theory and policy, an adequate solution to the problem would be a division of Czech state pensions into two pension pillars: a flat-rate pension and a social pension insurance scheme. In response to the ruling of the Constitutional Court, the Czech government did not opt for a radical pension reform, but for a change of several parameters of the existing “pension insurance” scheme. This is how the “small” pension reform was born, in which reduction thresholds and related reduction coefficients were re-established; these parameters were newly determined to reinforce the principle of equivalence between pensions and earnings to a certain extent. The government did so reluctantly (a “forced step”), because the opposite policy was closer to it – a gradual transition to a flat-rate pension, e.g. at the level of CZK 6,500. However, it did not do anything for that, even not later. From the perspective of the pension theory and policy, it is fully legitimate to propose (and enforce by means of the public choice) a flat-rate state pension as the only state pension system, or as one of two public pension pillars. The small pension reform de facto cancelled the second reduction threshold, or more precisely increased this reduction threshold many times: from 109% of NAE in 2010 to 400% in 2015, and in addition to that, it was merged with the earnings cap of 400% of NAE. The keyparameter for dependence of pension on earnings is thus onlythe first reduction threshold which newly has been even parametrised to 44% of NAE. At calculation of the percentage assessment of old-age pension, included are (only) 26% of earnings above this threshold, 100% of those below this threshold. The first reduction threshold with these two low parameters continues to be the key fact for characterising the Czech “pension insurance” scheme as a predominantly non-insurance system of Beveridgean type. After the small pension reform, the percentage pension assessment is slightly more transparent and comprehensible – thanks to an effective elimination of one reduction threshold. However, no radical reform occurred in this respect; the decision of the Constitutional Court was not strictly respected. The Czech basic assessment at the rate of 9% of NAE is the world’s rarity. We know its origin – from the compensatory contribution; however, for almost 20 years it has been existing (only) by force of habit. The amount corresponding to 9% of NAE is significant for old-age pensioners; nevertheless, it is not high enough for their subsistence. From this point of view, the basic assessment should be either increased many times or cancelled – depending on the public choice. The basic pension assessment cannot be compared to basic pension amounts in former labour systems of pension insurance schemes. The substantial difference among them is that at the pension calculation under the labour insurance scheme the basic amount (a given sum) was added to the “percentage pension sum” (earnings-related, without reductions, of course) to the total, further undividable pension sum; the assessed pension was (potentially) indexed as a whole. Nowadays, we could also cancel the very existence of the basic pension assessment in our country – a question can be even raised why we have not done it already many years ago, why we still regard this pension “assessment” as a compensatory allowance – 20 years after 1996! The compensatory allowance should have been one way or the other integrated in the pension amount in 1996 already, e.g. it could have been reflected in the reduction thresholds and coefficients. Even today, such cancellation of the very existence of the basic assessment would be a “mere” parametric reform of the existing defined benefit system. It would also lead to simplification of indexation of Czech state pensions which would become understandable. An alternative to dissolving the basic assessment of old-age pensions into a half-Beveridgean “percentage assessment” of old-age pensions is its significant increase to a level of a reasonable flat-rate old-age pension. Such a parametric reform is presently relatively easy to implement in our country – in that process the pension calculation for earnings exceeding the first reduction threshold can be used as a basis: the percentage assessment rate of 1.5% per year of insurance is to be multiplied by the reduction rate of 26% and we get: 1.5 * 0.26% = 0.39%. If this 0.39% rate is also applied on earnings lower than the first reduction threshold, what comes out is that we must increase the basic assessment to ca. 31% of NAE (see Figure 1). The raised problem of a reasonable adjustment of the basic assessment and of the percentage assessment of old-age pensions is thus resolved via a simple parametric reform during which we increased the basic assessment substantially and removed the reduction thresholds. This technical reform would be beneficial to low-pension recipients (or more precisely the scheme participants with their individual assessment basis not exceeding 44% of NAE). The old-age pension eligibility criteria would remain unchanged. Advocates of “respectable” pensions would have to be satisfied. Under the conditions of the year 2017, 31% of NAE represent almost 80% of the threshold for poverty at risk for an individual adult person; for two old-age pensioners we already get at ca. 105% of the same threshold for couples. In this variant of rationalising the calculation of new old-age pension, the quantified percentage pension assessment would be “definitively” added to the current basic assessment of the new old-age pension. And we would further index the total old-age pension only, as in the previous altervative. Figure 1 – Possible simple rationalization of pension calculation. Source: Author A standard component of most systems of protection of elderly in OECD countries is special, means-tested benefits for pensioners. A certain, typical exception is represented by countries with a Christian-democratic pension system that use segmented social pension insurance. Such a system puts a major emphasis on the principle of equivalence (including non-contributory insurance periods which can be covered by the state budget), whereas it lacks the solidary pension pillar. This is the case of e.g. Germany except that the social assistance system includes a special, meanstested (income- and property-tested) benefit for old-age and disabled pensioners (“Basic protection at old age and at reduced working capacity”) which consists of not only an allowance for subsistence, but also for housing, for payments of health insurance contributions or a potential allowance for increased costs, etc. The German “basic protection” (Grundsicherung) is basically provided for a period of one year. In their programmes, left-wing parties call for reinforcement of solidarity in the existing German pension system e.g. by introducing a minimum pension. Important theoreticians have made a conclusion that solidarity must be strengthened substantially – in the future, taking into consideration the German population ageing; e.g. Fehr et al. (2011) concluded in their econometric analysis that considerable reinforcement of progressivity of German public pensions is necessary: in their opinion, a flat-rate pension should represent 40% of public pensions. Typical liberal and social-democratic pension systems use housing benefits for senior citizens as an important additional pension benefit which is paid together with pensions by pension institutions. The social-democratic senior housing benefits are only income-tested – and they are utilised widely. Their importance to poorer citizens is great. In three Scandinavian countries, the basic pension benefit tested (for pension from social pension insurance) is a “guarantee pension” which guarantees the minimum pension, e.g. in Sweden at the level of approximately 24% of gross NAE (OECD, 2015), or one third of net NAE. The special housing benefit for senior citizens in Sweden is basically of the same design as the guarantee pension and is also administered by the state pension insurance institution. This benefit covers up to 93% of housing costs for single pensioners, specifically up to 15% of NAE (OECD, 2015). The housing benefit is not subject to tax, unlike the guarantee pension; the maximum amount of both these benefits in Sweden represents ca. 48% of net NAE. The guarantee pension is indexed to the price index, so its relative significance is expected to decrease. The Czech minimum old-age pension is defined as a sum of the basic pension assessment and the minimum percentage pension assessment. The latter has not been indexed throughout its existence yet (since 1996) and it continues to 0% 20% 40% 60% 80% 100% 120% 0% 50% 100% 150% 200% 250% 300% 350% 400% 450% Netpension(%NAE) Gross wage (% NAE) 2017 Reform amount to CZK 770 which currently represents 2.7 % of NAE, or. 3.6% of net NAE. The maximum amount of the housing benefit for a single pensioner in a town with population of 50 – 100 thousand is CZK 4,799 which is at the level of 17% of NAE, or 22.3 % of net NAE. In this standard example, the maximum housing benefit (17% of NAE) thus exceeds the minimum old-age pension (11.7% of NAE), which does not seem to be a rational relation. We can add the maximum housing benefit and the subsistence benefit at the level of the subsistence wage, i.e. CZK 3,410 per month (12.1% of NAE) and we get to the total of 29.1 % of NAE which is already only little less than 2% of NAE lower than the minimum amount of the “equalised” calculated old-age pension at the level of 31% of NAE. It is therefore evident that if we do not “cheat” old-age pensioners with earnings under 44% of NAE by means of a mirror image of the 26% reduction coefficient (in case of earnings exceeding 44% of NAE), then we would give them a total “reformed” pension as per Figure 1 and we would not have to provide them subsistence and housing benefits. Under “normal” conditions in the economic development, indexation of the basic pension assessment based on the wage index annually reinforces the role of the basic assessment in the total pension. In relation to that, an increase can be seen in the theoretical value of the minimum pension in the style of the reform according to Figure 1: the result we got in 2015 was 28.5 % of NAE, this year already 31% of NAE. The postponing of a proper reform results in an everincreasing strengthening of (already prevailing) solidarity over equivalence in our public pensions. The Czech public pensions require a radical reform – to separate the insurance pillar and the solidary pillar. The persisting division of pensions into the basic pension assessment and the percentage pension assessment does not definitely correspond to relative importance of the solidary and insurance components, especially because of the reduction of individual assessment bases exceeding (low) 44% of NAE to mere 26%. Indexation of the basic pension assessment according to the wage index further reinforces the role of this component and consequently increases the theoretical rational value of the flat-rate pension which was as high as 31% of gross NAE, or 40.7% of net NAE in 2017. 3 Reform Alternatives A paradigm pension reform is primarily a question of the public choice of one of four essential social models (Vostatek, 2016). From recent years we have experience with a neo-liberal approach to the existing Czech system consisting in the fact that the most important aspect is (although only partial) privatisation of public pensions, when the existing public pension system, non-standard, non-transparent and incomprehensible, is actually a current advantage, because the recommended introduction of a (mandatory if possible) private pension saving scheme brings, on the contrary, a transparent system, (technically) based on the principle of equivalence. In a longer term, some neoliberals (e.g. Kalousek) expected the existing public pensions to be replaced by the flat-rate pension of about CZK 6,500 which represented 26-27% of NAE in 2011–2013. At the present situation, we should or may expect from liberals and social democrats that they will support the transition to a flat-rate pension of approximately 30% of gross NAE, or 40% of net NAE. Advocates of the social-democratic model should or could push for complementing this flat-rate pension with as high as possible insurance pension; liberals would definitely not mind the insurance pension as a complementing pension pillar. It should be the biggest “problem” for Christian democrats, because “their” general pension welfare model is based on dominance of the social pension insurance – which is far from the current Czech system. However, they also could agree, to start off, with a paradigm pension reform which basically does not change the present relation of solidarity and equivalence in the public pension system; i.e. with transition to the flat-rate pension of approximately 30% of NAE and a separate insurance pillar which would fill the rest of the existing fiscal room. After the parliamentary election in autumn 2017, the coalition parties could thus agree on the paradigm pension reform relatively easily by means of which the system of two pension pillars would be introduced. Which of the two pillars mentioned will be strengthened or weakened in the future is a matter of further agreement. Rationalisation (and other) pension reforms also include indexation of solidary and insurance pensions. Here, international experience is undoubtedly important and interesting. However, it is not true that every pension social model could be matched with a certain system of indexations. The construct of increasing the already paid old-age pensions should or could be a follow-up of a global concept and a construct of a given pension pillar. What suggests itself in the solidary pension pillar is the link to the national average earnings (NAE) or to another indicator of standard of living. In the pillar of a social pension insurance type a lot depends on the initial calculation of the benefit, e.g. the original basis in the defined benefit system was “final” earnings, whereas the tendency nowadays is to calculate an old-age pension from life-long earnings – when former earnings are usually adjusted according to the wage level current at the time of the pension calculation (or more precisely the calendar year preceding the award of the pension). In notional defined contribution (NDC) schemes and in point systems, the valorisation of personal accounts and/or earnings, may be lower than according to the wage index, which creates room for a higher indexation of already paid pensions. “At the risk of oversimplification“, the authors of a special study on pension indexation (Piggott and Sane, 2009) concluded that • second tier schemes that aim for income-replacement (social insurance systems) could be indexed to prices, • first tier or poverty replacement schemes (solidary systems) should be indexed to community standards that include measures such as wages or per-capita GDP. These concepts of indexation prevail in OECD countries; nevertheless, they as such cannot be adopted simply. After all, the U.S. public pensions, which are significantly of the solidary type, are automatically increased by the index of pensioners’ costs of living. (The earnings valorisation uses the wage index.) Every author emphasizes that pension indexation is only one of the components of a given pension system. This also applies to reforms, no matter whether parametric or paradigmatic ones, even here, pension indexation should “fit” in the reform concept. An example of a reform designed this way is Great Britain, where during the transition to the new state pension, in 2011–2016, the existing basic state pension (BSP) was annually indexed to the highest of three figures: wage growth, price growth, 2.5%. The new British state pension, provided since 6 April 2016 (new fiscal year), is at the level of 21% of NAE. The British reform was paradigmatic – the state insurance pillar (State Second Pension) was gradually cancelled and simultaneously, the flat-rate BSP increased (from 16% of NAE, or 20% of net NAE, to 21% of NAE, or 26–27% of net NAE), as well as the occupational pillar supported by the state-owned pension company NEST grew step by step and the state support was cancelled for personal pensions, while the complementary solidary pillar (Pension Credit) was reduced. We can thus see in the example of Great Britain that pension indexation can be used for strengthening or weakening the flat-rate pension; generally, it is therefore not necessary to increase the flat-rate pension (or our Czech basic pension assessment) based on the wage index. The purpose of pension indexation is not to replace or extend the solidary pension pillar or its means-tested supplement. Equally, its purpose is not to replace the housing benefit for senior citizens or general welfare benefits. Perhaps the worst variant is to try to replace these benefits by a substantial minimum pension in the system of the (Czech) “pension insurance”, as required by the Council of Elderly Citizens of the Czech Republic (Rada seniorů ČR). For our elderly citizens, a comprehensible benefit would probably be a combination of social assistance and housing benefits according to the German model (Grundsicherung). A finding of the Council of Elderly Citizens significant for the national pension policy is that the existing housing benefits are used by elderly people only minimally although a considerable number of old-age pensioners are clearly entitled to them. Here, our public administration makes two mistakes: it does not search for citizens who are entitled to the housing benefit – moreover, this benefit is administered by the Employment Office of the Czech Republic (Úřad práce ČR)! – and it has not tabled the government a motion to reconstruct the housing benefit for senior citizens to make the benefit (at least) comprehensible to them and feasible. And in addition to that, pensioners should not be ashamed to ask for the benefit. Therefore, the minimum variant of the indexation reform is actually a reform of the housing benefit for elderly people according to the Swedish or German model on the understanding that the benefit will be administered and paid by the Czech Social Security Administration (Česká správa sociálního zabezpečení) as a provider of the existing “pension insurance”. Although it is not a change of indexation, this reform of the related housing benefit would probably reduce the pressure as to increase the indexation of low pensions and introduce a higher or high minimum pension. A minimum rationalisation reform of Czech public pension indexation itself is separating indexation of the percentage pension assessment from indexation of the basic pension assessment. If political parties do not reach the conclusion on necessity of a more important pension reform, then they should at least agree with introduction of indexation of the percentage pension assessment independently of indexation of the basic pension assessment. Maximum indexation would be indexation of the percentage pension assessment according to the dynamics of nominal NAE, whereas minimum indexation would be multiplication by the consumer price index and the “middle way” would be so-called Swiss indexation which uses the price index as one half and the wage index as the other half. The same generally applies to the basic assessment and there is no need to select the same index as in case of the percentage pension assessment. Consequently, it would result in such indexation that would be easy to understand by at least every other senior citizen. Of course, smart guys could start asking why we have two assessments of one pension – especially when each of the assessments would be indexed differently. I would not wonder if e.g. the current Premier would ask the current Expert Commission for Pension Reform this question. An already “bigger” pension reform in the Czech context would be a return to the historical system which was working under the labour social insurance for a long time: at the pension assessment, both pension “assessments” were added up and further recorded and indexed as a whole. When I proposed it at the meeting of the Commission, the chairman rejected this proposal promptly with justification that such a reform goes beyond the mandate of the Commission (Mandate, 2015) and thus he snuffed off the discussion. But it would be a miniature pension reform which would not touch the core of the existing system at all; it would only open the way for the introduction of simple and understandable indexation of pensions. Nevertheless, it is possible that experts or journalists or even amateurs would then start wondering e.g. why we have the basic pension assessment – when we are able to do without it (like in the USA) – and why this assessment is considerably lower than any other flat-rate pension in the world. But they can (and even should) ask this question already now. No opinion in this respect was given by Potůček’s Commission. The political expert and the vice-chairman of the trade union confederation ČMKOS Samek even says repeatedly that we do not need any reform of pension calculation, because it would bring nothing to people; that if people do not understand the calculations (including indexation), it does not matter at all – the state will calculate their pensions for them. Theoretically, a step forward would consist in a slightly more distinct pension reform: splitting the current “pension insurance” into two pension pillars, with separate indexations. However, the basic assessment (newly the flat-rate pension) would be too small and approximately half of the percentage assessment (further called the pension insurance) would be a solidary pillar. This reform would make sense only if considerable changes of parameters were further reckoned with – they would e.g. mean an increase of the flat-rate pension and a decrease of solidarity in the pension insurance. It is theoretically possible, but international experience shows that a rather radical pension reform cannot be implemented via multiple gradual steps. Much more feasible is to make a paradigm reform. This applies to Czechia all the more so, because here it is possible to make a modern, radical pension reform in which no change would occur in the overall ratio of solidarity and equivalence in new public pension entitlements; the pensions already being paid now would not in fact change at all – they would only be “re-defined” according to the split of the total pension into the new flat-rate pension and the “remaining” insurance pension which would be quantified by subtracting the flat-rate pension from the existing total pension entitlement. Both pensions would then have separate indexations. If the new insurance system was the NDC system, it would substantially facilitate further potential pension reforms – if e.g. some of the governments wished an increase the pension under this system, it would suffice to increase the insurance premiums to this system – and this would be reflected in the future final pension savings (entitlements) on the client’s personal account in the NDC system which turn into the life-long pension at the pension assessment: by multiplication by the corresponding conversion rate. At a one-time transition to the NDC all the existing pension entitlements of the insured turn into a deposit on their personal account. The whole NDC system is transparent and comprehensible – and this also applies to the pension indexation systems. An expert answer to the Premier’s requirement to look for a more balanced method of pension indexation is: „Progressive indexation – when effective rather than “tokenist” – is a poor and arbitrary substitute for a progressive benefit formula or a two-tier pension system with redistributive and earnings-related elements“ (Whitehouse, 2009). Indexation cannot compensate for “insufficiently respectable” amount of the newly assessed old-age pensions. “Another way to think about the growth of benefits is to recognize that, at a given long-run cost, there is a trade-off between the initial benefit and the rate of growth of benefits thereafter: the more rapidly benefits grow, the lower the initial replacement rate must be to hold cost constant” (Barr and Diamond, 2008). 4 Conclusion The Czech “pension insurance” is a non-transparent and incomprehensible mix of a low flat-rate pension and semi-insurance pension which is largely deformed by a low reduction threshold (44% of NAE) and an extra low reduction coefficient (26%). Combining these two deformative parameters practically turns against low-income participants of the system. The paper shows how it is possible – in this highly specific situation – to implement a formally paradigmatic, yet practically only technical pension reform which would result in a many times higher flatrate pension or the basic pension assessment (ca. 31% of NAE) and a typically insurance-based pension, without changing the total amount of pensions calculated from earnings exceeding 44% of NAE. Who would “profit” from this reform are only those participants with average earnings of less than 44% of NAE who typically do not use the housing benefit nowadays, because it is too complicated for them also due to the fact that they would have to document their housing costs to the employment office quarterly; moreover, the public administration does not search for such elderly people. Therefore, what would be beneficial for low-income senior citizens is the very reform of the housing benefit in which a change would be made in providing the documentation for the entitlement and assessing the benefit once a year and the administration would be done by the pension provider. The highly complicated construction of the entire system of “pension insurance” and of the housing benefit is a source of misunderstanding in the pension indexation process which as such is also incomprehensible: mainly due to the complicated formula for calculation of indexation of the percentage pension assessment in which indexation of the basic pension assessment is deducted. One of the consequences of such a highly complicated situation is a demand for a “more balanced” method of pension indexation by the Premier, aiming to further equalisation of pensions. The purpose of pension indexation is not to correct the deficiencies in the basic concept of pensions or housing benefits. However, pension indexation can be used for a (gradual) reform of the pension system as it was the case of e.g. Great Britain. In Czechia, pension indexation could be used for a subsequent modification of parameters after the above-mentioned technical reform of pensions is implemented. Indexation of the current Czech pensions can be simplified and made understandable even without any link to an overall pension reform, only by a mere summation of the basic and percentage pension assessments at their calculation and subsequent indexation of the total pension assessment. Acknowledgement The paper was prepared as part of solving the student project “Improvement of the pension system as an intensification factor of the economy”, using a specific grant for a specific academic research of University of Finance and Administration. References [1] BARR, N., DIAMOND, P. (2008). Reforming Pensions: Principles and Policy Choices. Oxford, Oxford University Press [2] CONSTITUTIONAL COURT (2010). Rozhodnutí Ústavního soudu ČR: Nález ÚS 8/07. http://www.asocr.cz/addons/files/stare/110107rozhodnuti.pdf [3] FEHR, H., KALLWEIT, M., KINDERMANN, F. (2011). Should pensions be progressive? Yes, at least in Germany! CESIFO Working Paper No. 3636. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1961052 [4] MANDATE (2015): Mandát Odborné komise pro důchodovou reformu. http://www.duchodova- komise.cz/?page_id=14 [5] MF (2012). Důchodový účet nebyl zrušen, protože neexistoval. Polemika s článkem v Haló novinách. Praha: tiskové oddělení ministerstva financí. http://www.mfcr.cz/cs/aktualne/v-mediich/2012/2012-07-11- vmediich-4914-4914 [6] OECD (2015), Pensions at a Glance 2015: OECD and G20 indicators. Paris, OECD Publishing. http://dx.doi.org/10.1787/pension_glance-2015-en [7] OKDR (2017). Hledání vyváženějšího způsobu valorizace důchodů. Praha: MPSV. http://www.duchodova- komise.cz/wp-content/uploads/2017/06/Hledání-vyváženějšího-způsobu-valorizace-19.-dubna-2017.pdf [8] PIGGOTT, J. and SANE, R. (2009): Indexing Pensions. Social Protection Discussion Paper No. 925. Washington: The World Bank. http://documents.worldbank.org/curated/en/686271468155723487/pdf/524450NWP0Box345558B01PUBL IC100925.pdf [9] VOSTATEK, J. (2016): Penzijní teorie a politika. (Pension Theory and Policy). Prague, C. H. Beck [10]WHITEHOUSE, E. R. (2009). Pensions, Purchasing-Power Risk, Inflation and Indexation. OECD Social, Employment and Migration Working Paper No. 77. OECD Publishing. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.463.4211&rep=rep1&type=pdf Contact information Prof. Ing. Jaroslav Vostatek, CSc. University of Finance and Administration Address: Prague 10. Estonská 500 E-mail: jaroslav.vostatek@vsfs.cz