Scholarly article on topic 'On the Sustainability of the Romanian Pension System in the Light of Population Declining'

On the Sustainability of the Romanian Pension System in the Light of Population Declining Academic research paper on "Economics and business"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"population projection" / "Lee Carter model" / "social security" / "old age pension" / "dependency ratio"}

Abstract of research paper on Economics and business, author of scientific article — Ciprian Pânzaru

Abstract The paper is focused on the analysis of the demographic risks which affect the social security system. The situation is analyzed based on projections of the population from Romania, for the period of time 2010 – 2060. The population projection was made by an adaptation of a cohort component method. For the projection of the mortality rate was used the Lee-Carter methods. The same method was adapted for the projection of the fertility rates. For the projection of migration has been used an autoregressive integrated moving average model. Based on the data obtained as result of the population projection have been imagined some scenarios, regarding the sustainability of the social security system. These scenarios analyzed the way in which the sustainability of the pension system could be ensure, starting from the regulating role, which the migration may have on the labour market. The results show that the migration is the only solution to sort out the labour market deficit during the next 50 years. According to estimations, the Romanian labour force should be supplemented till 2060 yearly with 200,000 to 500,000 immigrants.

Academic research paper on topic "On the Sustainability of the Romanian Pension System in the Light of Population Declining"

Available online at www.sciencedirect.com

ScienceDirect

Procedia - Social and Behavioral Sciences 183 (2015) 77 - 84

International Workshop on Ideologies, Values and Political Behaviors in Central and Eastern Europe

On the Sustainability of the Romanian Pension System in the Light

of Population Declining

Ciprian Pânzaru *

West University of Timisoara, Department of Sociology, Timisoara 300223 Romania

Abstract

The paper is focused on the analysis of the demographic risks which affect the social security system. The situation is analyzed based on projections of the population from Romania, for the period of time 2010 - 2060. The population projection was made by an adaptation of a cohort component method. For the projection of the mortality rate was used the Lee-Carter methods. The same method was adapted for the projection of the fertility rates. For the projection of migration has been used an autoregressive integrated moving average model. Based on the data obtained as result of the population projection have been imagined some scenarios, regarding the sustainability of the social security system. These scenarios analyzed the way in which the sustainability of the pension system could be ensure, starting from the regulating role, which the migration may have on the labour market. The results show that the migration is the only solution to sort out the labour market deficit during the next 50 years. According to estimations, the Romanian labour force should be supplemented till 2060 yearly with 200,000 to 500,000 immigrants.

© 2015TheAuthors.PublishedbyElsevier Ltd.This is an open access article under the CC BY-NC-ND license

(http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer review under the responsibility of the West University of Timisoara.

Keywords: population projection, Lee Carter model, social security, old age pension, dependency ratio

1. Introduction

Worldwide, the discussions regarding the sustainability of the social security systems are of highest actuality, moreover because, under the nowadays socio-economical circumstances, these undergo an acute crisis. Thus, the factors like the population aging, the diminishing of the active population, the migration of the labour force etc.

* Corresponding author. Tel.: +40-256-592-148. E-mail address: ciprian.panzaru@e-uvt.ro

1877-0428 © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer review under the responsibility of the West University of Timisoara.

doi: 10.1016/j.sbspro.2015.04.848

represent major challenges for a configuration of the politics regarding the social security systems. The importance of this aspect is reflected by the allocation from GDP, in most European countries, which are between 8 and 11%, in order to ensure a minimum stability of the social security systems.

The nowadays situation is characterized in Romania by a profound transformation of the economical structures, a diminishing of the population, an increase of the attending of youth to the higher education level, and especially by massive pensioning of the population and by migration. All these generated a drastic diminishing of the presence on the labour market.

Under these circumstances, a major social risk is generated by the growing of the non-contributing population segment, from the whole active population. This situation is due to several factors. On one hand, we talk about of a demographic dynamics, characterized by a diminishing of the birth rate, which generates a population decrease and by an enhancing of the life expectancy, which generates a population ageing - and on the other hand, we see a higher migration, which cooperates in creating of a general misbalance between the active and the inactive population. The existence of a major segment of economically active population, eager to emigrate, affects the taxation basis, the incomes to the State budget and especially the sustainability of the social security systems. Due to these causes, on the labour market will appear, gradually, less numerous generations, and the amount of employees will not grow extremely much, even in the event of a constant economical growth, in the inexistence of a demographical basin, firm enough for a growth of the mass of employees. As result, the supplies of young labour force will be more and more diminished, being also affected by the growth of the percentage of people attending higher education forms from each generation, as well as by the migration, which is supposed to enhance, especially if the crisis will continue and at the same time, with the liberalization of the labour market in the EU states. In one more generation, starting with 2030, on the labour market will come the children born after 1990, who are even fewer, the deficit of labour being thus emphasized. The effects of these demographic waves were, are and will be sources of major misbalances in the educational systems, on the labour market, in the social security system, in the migration processes, and as well in the development of the structure of the Romanian population. The reducing of the mass of employees and, moreover, the growing of the mass of retired persons is generated, beyond the process of demographic ageing, also by a substantial growth of the life expectancy. Therefore, the proportion between the persons of productive age (15 -64 years) and the persons with an age over 65 years will deepen constantly.

This demographic development has repercussions on the whole social security system. An appreciable amount of inactive person means, in fact, a rising of the amount of beneficiaries of the social security system. The burden falls on the pension system, especially due to its actual structure pattern, mainly based on the contributions principle, and intergenerational solidarity.

2. Methodoloy

The research started from the premise, that the population is a relationship of the type y = f (x) , where the response y of the system is a function of three process variables, namely: the death rate, the fertility rate and the migration. The knowledge, the forecast and the estimation of these three variables may offer us a coherent image upon the mood in which, from the point of view of its sustainability, the social security system will look like, in future.

In order to assess the death rate, we used the Lee Carter model (Carter & Lee, 1992). The model was elaborated by Ronald D. Lee and Lawrence Carter in 1992, being presented for the first time in the essay „Modeling and Forecasting the Time Series of U.S. Mortality", published in the Journal of the American Statistical Association. We also used an adaptation of this model for the assessment of the fertility rate (Panzaru, 2014).

The Lee Carter model uses specific death rates for certain age groups, in order to estimate, a mortality index ( kt ),

(or a fertility index, ft, in our case). This index is calculated as well for the whole population, as also for each

gender group: male and female. Once the index is estimated, we may predict the death rate and the life expectancy. The model starts from the premise, that between the logarithmical value of the specific death rates and other two variables: age and time, there is a linear relationship.

The model proposed by Lee Carter may be expressed by the following formula:

ln(mx,,) = ax + bxkt +Sx,, (1)

where ax is the average age-specific mortality; bx is a deviation in mortality, due to changes in the kt, kt is the

mortality index in the year t and Sx is the error term with mean zero and variance G E , reflects the age-time

variation, not captured by the model.

The model shows the fact, that for each moment of time and for each age, we will have various values of the death rate. Therefore, we may talk about a simultaneous system of equations to be solved, in order that we may assess the values for ax, bx and kt. In order to obtain for this system a unique solution, we have to impose two

constraints: the sum of the parameters bx = 1, and the sum of the parameters kt = 0 :

xmax tmax

Z bX = 1 and £ kt = 0 (2)

X=xmin t=tmil

Under the circumstances of this assumption, ax is the average values over time of the ln(mx t) for each X . Therefore, a x coefficients are calculated with the following expression:

Z ln(mx,t)

ax = —--(3)

In order to find out the value of the parameters bx and kt, the singular value decomposition method (SVD) was applied to the matrix M = |ln(mxt) — ax }. Once the parameters ax and bx are estimated, we can find kt (or

ft). Further, for the modeling of the kt (or ft), we used the autoregressive integrated moving average model -

ARIMA (Box and Jenkins, 1976; Hamilton, 1994). Finally, based on the previous calculated parameters, we estimated the specific death rates, on age and gender groups (or only on age groups, in case of the fertility rates). For migration projection we used an ARIMA model (autoregressive integrated moving average) (Panzaru 2014). Further, once the three indicators are estimated (death rate, fertility rate and migration), we used for the projection of Romania's population in the time period 2010 - 2060 the cohort component method (Whelpton, 1936, Klosterman 1990). The components are considered to be independent processes, which cause various changes, regarding their amplitude, and in separate moments of time, in the population segments. This method starts from the following idea: if at the t moment, we have a population with a certain structure, on ages and genders, the number of people at time t +1 is equal to the number at time t plus the number of births, minus the number of deaths, plus the number of immigrants, minus the number of emigrants

3. Results

Regarding the aspects which set-up the social security system (by its main sub-components) and the population, the debate starts from the diminishing of the mass of tax payers. In theory, this relationship may be translated by the dependency ratio. The dependency ratio represents the quotient between the population with ages between 15 and 64 years, and the population with an age over 65 years and the one with an age under 15 years. Upon its value intervene the three factors, which shape the population amount: death rate, fertility and migration. The population with ages from 15 to 64 years represents a potential contributory population, regarding the social security systems (pension, health system, unemployment).

A low fertility rate leads to the diminishing, in time, of the population effectives, while the death rate variation, on various age groups, may generate misbalances within the population, especially regarding the ratio between contributors and beneficiaries. For example, 50 years ago, as the life expectancy was in Romania, of about 68 years,

and the retirement age was set at 62 years, we talk about a certain sustainability of the pension system, totally different of the present situation, when the life expectancy rose with over 10 years. At the same time we have the migration, which causes substantial changes in the financing of the social security systems. The migration and specifically the emigration affects nowadays the population amount, and especially the economically active population.

The present paper is based on an own population projection for the time period 2010 - 2060, starting from separate estimations of the three variables: death rate, fertility and migration. In figure 1 is presented the projected population development, for the time period 2010 - 2060:

Fig. 1. Romania's projected population for the time period 2010 - 2060.

This projection joins the ones of the United Nations Organization (13.9 million inhabitants in 2080), of Eurostat (16.3 million inhabitants in 2080) and of the National Statistics Institute, which is the most recent projection (14.27 million in 2060). The present projection estimates a volume of 14.57 million inhabitants in 2060.

As we may see, Romania's population is in constant diminishing. The population diminishing leads undoubtedly also to a diminishing of the economically active population, and to a considerably reducing of the attendance on the labour market. This sure causes also effects on the social security system, the sustainability of which is mainly based on the principle of solidarity between generations. The sustainability of the social security system is given by an optimal ratio between contributors and beneficiaries. Accordingly, the basic element on which is based a social security system is the population, generally speaking, and the employed population, especially. A population in diminishing will finally generate also a diminishing of the employed population, as well as an increasing population will lead to an increase of the employed population. Therefore is the most important issue in the debate on the social security systems and, more specific, on their sustainability, the population and its evolution.

In table 1 are presented the population volumes, on the main age groups brought forward in the present research:

Year 0-14 15-64 65+ TOTAL

2010 3093009 13935006 3343335 20371350

2015 3059746 13629820 3536934 20226500

2020 2968737 13042860 3923118 19934715

2025 2760568 12515826 4251948 19528342

2030 2524809 12238886 4262038 19025733

2035 2374429 11526924 4555535 18456888

2040 2256221 10804433 4733026 17793680

2045 2116845 10060211 4862527 17039583

2050 1955111 9507011 4781441 16243563

2055 1799640 8823193 4810963 15433796

2060 1662650 8321250 4594462 14578362

Table 1. Romania's population for the time interval 2010 - 2060, main cohorts.

We notice in this table a constant diminishing of the population from the age segment 15-64 years, simultaneous with an increase of the population of over 65 years of age. Clearly is under these circumstances, the most one affected the social security system, and especially the pensions system. The pensions costs represent the biggest category of budget charges, namely, one fifth of the public expenses.

4. Discussion

Due to the diminishing of the population will enter, gradually, less numerous generations on the labour market, and the amount of employees won't rise very much, even in the case of a constant economical growth, in the absence of a strong enough demographic basin, in order to be able to enhance the amount of employees. The solution is to increase the number of contributors. But this segment is closely linked to the total amount of population. As the latter diminishes, it is expected that the tax paying population will be also reduced. On the other hand, due to the ageing of the population the segment 65+ will register a considerable growth. The ratio between the individuals of economically active age (15 - 64 years) and the people with an age over 65 will deepen constantly.

If this premises will continue, then the dependency ratio (the ratio between the persons with an age under 15 years and the one over 65 years on the one hand, and the persons with an age between 15 and 65 years on the other hand) will start from 0.46 in 2010 and will finally reach up to 0.75.

Yet does not all population with an age between 15 - 64 years represent also a population which contributes by tax paying to the social protection and security systems. Therefore is more relevant the analysis of the economic dependency ratio (the ratio between the inactive population - of under 15 years, of 15 years and over, the unemployed - and the active/employed population). For this ratio has been considered a constant unemployment rate of 7%. Under these circumstances, starts the economic dependency ratio from 1.43 in 2010 and may reach up to 1.91 in 2060. This means that to a single working person will be allotted almost two inactive persons.

In its efforts to reduce even more the unemployment expenses, the European Council established 2010 in Lisbon that, based on a lasting economical growth of 3% of the GDP, the member states should direct their steps to a total average rate of the labour occupying of 70%, as well as over 60% for women. If in 2020, we would have an employment rate of 70%, the dependency ratio would diminish to 1.18 for that year, but it would afterwards resume its ascendant trend immediately during the next period, getting up to 1.50 in 2060. So that we notice that, in case of Romania, even under circumstances of an increase of the occupation rate, the economic dependency ratio will worsen.

Another scenario takes into consideration the increasing of the retirement age from 2035 on, to 70 years, as well for men, as also for women. The increasing of the retirement age is the solution seen by many, in order to use as much as possible the resource of the economically active population. Nevertheless represents the mobilization of the inactive human resources the main priority for the next years, due to the fact that most EU member states (including Romania) have a still unused labour force capacity, which can be activated during the next 10-15 years. But the maintaining of the elder persons on the labour market does represent a problem, due to the fact that the retirement before the legal retirement age has become, especially for Romania, an usual practice during the past 2 decades. In

case of this scenario was maintained the condition of an employment rate of 70% starting with 2020. Under these circumstances, the economic dependency relation (starting with 2035) was calculated as quotient between the inactive population - under 15 years, over 70, or unemployed - and the working population. Under these circumstances, the economic dependency ratio starts from the same value, of 1.43 in 2010, diminishing to 1.25 in the year 2060. But it has to be emphasized that this diminishing is only apparent. In fact would the dependency ratio diminish only from the moment of initiation of this measure on (reaching on 2035 the level of 1.05), but would start increasing afterwards.

5. Conclusions

It seems that none of the upper imagined sceneries does guarantees a sustainable pension system. Not even if we would increase the employment rate to 100% (which is totally unlikely), would be the sustainability of the pension system be ensured, and till 2060 the dependency ratio would still follow an ascendant trend.

But these data represent just the top of the iceberg. We must still take into consideration that not all population of employment age (15 - 64 years) is formed by individuals paying taxes to the social security system. The population able to work is not the same to the employed population, as on the other hand, neither the employed population isn't the same to the amount of employees. The occupation rate was in Romania in 2010 of 58.8%. This figure is placed beyond the European average, of 64%. But in Romania appears a significant difference between the volume of the employed population and the volume of employees, which can be explained by the high amount of unremunerated workers (part of the employed population). In Romania is the share of non-remunerated family workers (meaning, of non-tax payers) of about 25% of the employed population (more exactly, of 26.8% in year 2010). Regarding the amount of retired persons, we had 2010 in Romania 4.7 million pensioners (Ministry of Work, Family and Social Security, 2010). Among these are 72.2% retired for age limit.

If we stick to the initial premise, according to which a percentage of 72.2% from the total amount of pensioners are pensioners for age limit, we could estimate that in the year 2060, we could have an amount of almost 6.5 million pensioners.

Therefore would the hypotheses from the previous sceneries, even if they would prove real in time, still not be sufficient for providing of a minimal sustainability for the pension system, and implicitly of the social security system. For example, if we could maintain the nowadays conditions (an occupation rate about the value of 60% and a retirement age of 65 years, as well for men as also for women), the ratio between contributions and expenses would evolve from 0.91 in 2010 to 0.40 in 2060. This means that only 40% of the pensioners would have the pensions ensured only from contributions of the employees. The ratio contributions-expenses was estimated by taking into consideration of a pension share of 35% from the average wages. The level of the contributions to the social security system was estimated as constant 31.30% (10.50 individual share, 20.80 employer's share)

Under these circumstances is the only solution represented by a completion of the economically active population with immigrants. But there rises the question „how much immigrants are needed, in order to fill up the labour deficit, in order to provide a proper sustainability for the pensions system?".

Taking as reference the ratio between contributions and expenses, in case of the first scenario, the mass of employees should increase with about 200,000 persons yearly until 2020 and should reach the necessary amount, of almost 1,000,000 persons yearly during 2055-2060, which is totally unlikely.

In exchange, under the conditions of an increase of the occupation rate to 70% starting with 2020, we would need a supplementation of the mass of employees with about 100,000 persons yearly until 2025, with about 200,000 persons yearly between 2025 and 2030, reaching to a necessary of about 800,000 persons yearly during the time interval 2055 - 2060.

The most likely situation is the one, in which the occupation rate would increase to 70% in 2020, cumulated with an increase of the retirement age to 70 years, as well for men as also for women, starting with 2035. This situation would necessitate only an increase of the employees with 500,000 until 2035. After that moment, we would necessitate an infusion of employees of about 200,000 persons yearly. We remark a growing of the necessary of employees in the time interval 2050 - 2060, when the necessary amount would be of about 500,000 employees yearly. Figure 2 shows the needed employees, in case of the three imagined sceneries:

5,000,000-

4.000,000"

3,000,000-

2,000,000-

1.000,000-

Fig. 2. The needed

Clear remains that only part of these may result from the active population (under the conditions of an increase of the employed population), but most of them will have to be offered by immigration. Beside the mobilization of the inactive resources, the migration might have a substantial contribution in the insurance of a balance on the labour market and in compensating of the demographic decline. Clearly is the imposed condition also a permeability of the labour market, in order to be able to absorb this resource. In its absence, the migration might have a negative effect, with misbalances especially on the social security system.

The migration role is well-known for solving of the misbalances from the labour market. In the assessments of the European Commission is estimated that the immigration „has contributed significantly to the increase of the occupation of man power in the European Union during the past years" (Platonova & Urso, 2010, p. 257). Thus, during the time period before the starting of the economical crisis (2000 - 2007), accordingly to the estimations of the European Commision, the migrants from third countries contributed to an increase of the employed population with almost 3.7 million persons, representing about one quarter of its total increase (idem, p. 260).

During 1994 - 2004, Spain created over 6 million jobs, from which a third were occupied by migrants. Norway created about 250,000 work places during 2000 - 2008, and almost half of them were occupied by migrants.

The migration is steadily influence the population structure, being sometimes a regulating factor of the misbalances, which may appear in the development of the population, and at other times a factor just to worsen this misbalance. For example, Germany has regulated its issue of the population diminished after the Second World War by mass acceptance of the immigrants from Turkey. So that the labour market did not suffer and the tax payers were in sufficient amount, to be able to provide the needed support to the pensions.

References

Box, G. E., & Jenkins, G. M. (1976). Time series analysis: forecasting and control, revised ed. San Francisco: Holden-Day. Eurostat. (2012). The 2012 Ageing Report: Economic and budgetary projections for the 27 EUMember States (2010-2060). Retrieved from http://ec.europa.eu/economy_finance/publications/european_economy/2012/2012-ageing-report_en.htm.

-1-1-1-1-1-1-1-1-1-1-1—

2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060

amount of employees during 2010 - 2060

Hamilton, J. D. (1994). Time series analysis (Vol. 2). Princeton: Princeton University Whelpton, P. K. (1936). An empirical method of calculating future populationJournal of the American Statistical Association, 31(195), 457-473.

Klosterman, R. E. (1990). Community analysis and planning techniques. Savage: Rowman & Littlefield Publishers.

Lee, R. D., Carter, L. R. (1992). Modeling and forecasting US mortality. Journal of the American statistical association, 57(419), 659-671.

Ministry of Labor, Family and Social Protection. (2014). Number of State Social Insurance Pensioners. Retrieved from http://www.mmuncii.ro/j33/index.php/en/services/statistics/statistic-data.

Panzaru, C. (2014). Effects of Labour Migration on Social Security System. Unpublished doctoral dissertation, West University of Timisoara, Timisoara, Romania.

Platonova, A. & Urso, G. (2010). Migration, employment and labour market integration policies in the European Union: Part 2. Brussels: International Organization for Migration.

Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat (2012). World Population Prospects: The 2012 Revision. Retrieved from http://esa.un.org/unpd/wpp/index.htm.

Romania's National Institute of Statistics. (2013). Projection of the population in Romania at the horizon oof2060. Bucharest: INS.