Assessing the Benefits Reform in Slovenia Using a Microsimulation Approach

Transcription

1 Assessing the Benefits Reform in Slovenia Using a Microsimulation Approach Nataša Kump Institute for Economic Research Kardeljeva pl. 17, 1000 Ljubljana natasa.kump@ier.si Tel: +386(0) Boris Majcen Institute for Economic Research Kardeljeva pl. 17, 1000 Ljubljana majcenb@ier.si Tel: +386(0) Mitja Čok Faculty of Economics, University of Ljubljana Kardeljeva pl.17, 1000 Ljubljana mitja.cok@ef.uni-lj.si Tel: +386(0) ABSTRACT: Slovenia faced a wave of proposed structural reforms in Consequently, the door to acquiring data at the individual level was finally opened and the database for a national tax microsimulation model was built. This microsimulation model soon proved to be a useful analytical tool for evaluating the consequences of particular policy measures in the area of social security. A new regulation governing social transfers is supposed to be enacted in June A second version of the Slovenian static microsimulation model has been constructed to assess the effects of this ongoing comprehensive reform. The model is based on a cross-sectional dataset representing a population of all ages at a certain point in time (the year 2007). The sample consists of 40,000 households with approximately 115,000 individuals (approximately 6% of the population) and is based on linked data taken from statistical and administrative sources. The paper describes the main characteristics of the current system of social s, reform changes and the structure of the model which was used to identify the winners and losers as well as the distributional effects of the reform. 1. INTRODUCTION A significant part of Slovenia s current social system was formed at the start of 1990, including the new Social Assistance Act and Family Benefits Act. Since then, some changes have been introduced to social s. At first, the child was just a kind of basic social protection (social assistance), while in 1998 the child became less of a social assistance disbursement and more of a family policy. In the period from 1993 to mid-2001, the guaranteed minimum income (social assistance) in Slovenia was very low. The Social Assistance and Services Act was amended in 2001 and higher levels have gradually been introduced. Due to the high child level for children in families with the lowest income, the 2001 amendments to the Social Assistance and Services Act set the minimum income for children at a low level (30% of the minimum income of the first adult in a family) (Stropnik and Stanovnik, 2002). Both s are disbursed by the country s centres for social work. The Social Assistance and Services Act also regulated subsidised rent until the Housing Act was adopted in Municipalities are responsible 1

2 for rent subsidies. A new Scholarship Act introducing a new means threshold for entitlement to state scholarships was adopted in The Act for the first time independently governs the scholarship scheme which has been regulated by the Employment and Insurance Against Unemployment Act prior to the change. With this new regulation, the responsibility for state scholarships was transferred from the Employment Service of Slovenia to the centres for social work. Although the pension system contains various redistribution elements, some pensioners still receive very low pensions due to their low incomes during their contribution period and/or an incomplete contribution period. Pensioners with low incomes and an incomplete contribution period can apply for a pension income supplement provided they fulfil a set of conditions, such as: a) their pension is below a certain threshold; b) the family in which the pensioner lives does not have other income sources which would allow a minimum standard of living; and c) the family has negligible property. The 1999 Pension and Disability Insurance Act (PDIA) introduced the state pension which is in fact a form of a means-tested. As a social assistance measure, it has no place in the pension and disability insurance act; its inclusion was the result of considerable horse-trading among the political parties making up the ruling coalition. Unlike the pension income supplement, where one of the main criteria is that a person is entitled to a pension, i.e. has accumulated a sufficient contributory period, the state pension can be granted to someone who is not receiving any pension. Both s are disbursed by the Institute for Pension and Disability Insurance. Subsidies form an important part of the Slovenian system of social protection. Childcare services are affordable due to the high means-tested subsidies from public sources. Regulated by the Pre-school Childcare Act, these subsidies are by far the biggest single transfer to families in Slovenia and are governed by municipalities. Children in compulsory and upper secondary education might be eligible for subsidised school meals with some of these entitlements being universal rights and others means-tested. Subsidised school transport is provided for pupils in upper secondary education and tertiary students. The Slovenian social protection system plays a significant role in maintaining low poverty rates in Slovenia. However, some drawbacks have become evident over time: a) Many state and municipality authorities are involved in assessing eligibility and disbursing the s. Official records are not completely harmonised among the different authorities which leads to the possibility of incomes being underreported. b) The Slovenian social protection system enables the accumulation of s: a) relating s to the fact that one is receiving some other ; b) the non-inclusion of all incomes and s in family income when deciding on entitlement; c) the 2

3 definition of income ceilings for social entitlement; d) the absence of a linkage among registers of beneficiaries of various social s; and e) deficient co-ordination and co-operation among authorities granting s etc. For example, the 10% childcare subsidy and free lunches in primary school that are granted exclusively to social assistance beneficiaries. c) Work incentives are low. d) The minimum income (which is the basis for social assistance and housing s) has become too low. Therefore, the regulation concerning social s has been amended and two new Acts have been adopted. 1 The Exercise of Rights to Public Funds Act introduces a single entry point and harmonises the criteria for granting four types of transfers and nine types of subsidies that are meanstested. The two main goals of the changes are to simplify the procedure of applying for transfers and subsidies for the beneficiaries, while making the system more transparent and efficient and reducing the possibility of fraud (by cross-checking income and wealth status). Centres for social work will become the single entry point in the system of social rights. Their work on applications will be supported by a centralised information system (a web application connecting different databases). The Financial Social Assistance Act introduces changes in the areas of financial social assistance and minimum pension support, previously governed by the Pension and Disability Insurance Act (National Reform Programme , 2011). For the purposes of assessing this ongoing comprehensive reform a second version of the Slovenian static microsimulation model has been constructed. The microsimulation model was a tool for estimating budgetary effects, identifying the winners and losers and assessing the distributional effects. The structure of the paper is as follows: Section 2 outlines the data and methodology, Section 3 briefly describes the changes in social s, the results are presented in Section 4, while the final section offers some concluding remarks. 2. DATA AND METHODOLOGY For the purposes of the reform simulations, previous versions of the microsimulation model (Čok 2002, Čok et al. 2004, Majcen et al. 2007) were updated and further developed. This second version of the Slovenian static microsimulation model is based on a cross-sectional dataset representing a population of all ages at a certain point in time (the year 2007). The sample consists of 40,000 1 The Acts were supposed to commence operation in June Since adequate records are not ensured (especially property valuation models) and the information system is not established yet the Acts will be enforced in January

4 households with 112,839 individuals (approximately 6% of the population). It is a merged database based on separate files from the Ministry of Finance (containing PIT records), the Ministry of Labour, Family and Social Affairs (containing records on social s), the Ministry of Internal Affairs (containing data on socio-economic characteristics), the Employment Service of Slovenia (containing data on unemployment s and labour market status) and the Ministry of Education and Sport (containing detailed data on scholarship ). Since the model is based on administrative data (not survey data) it is high in quality and enables the simulation of practically all direct taxes and most s at the individual and household levels. The model is static in time and does not include changes in behaviour or consequential macroeconomic second-order effects. In the first step of the microsimulation, databases from different sources are prepared in the final form (final database) which enables the simulations. Since many households consist of two or even more families (for example, adults with minor children, spouses and their parents) we imputed families within these households. At the end, our database consists of 112,839 individuals living in 57,630 families. Secondly, the system from 2007 is replicated to estimate the quality of the sample. Since the original data also include social s calculated by a responsible authority, one might expect the micro validation of the model could be done by comparing the s calculated by the model with the recorded social s for each individual in the sample. However, this cannot be done in such a straightforward way due to a few data limitations: Family income which is the basis for the income test for most s refers to the previous year or even the year before that, meaning that 2006 income data are needed to simulate social s in As income data refer to 2007, our assumption is that income levels in 2007 are exactly the same as in Social assistance is a special case since entitlement to social assistance depends on family incomes in the three previous months. Given the data on annual incomes it is implicitly assumed that income is received at the same rate throughout the year and monthly incomes are simply annual incomes divided by 12. The actual income position of individuals might drastically change during a single year and a three-month average could differ from the annual average. Social assistance is usually not received throughout the year, therefore we take into account that social assistance is received on average 7.45 months per year (official data). Entitlement to some social s (social assistance, pension income supplement) also depends on family savings and property. Data on savings and property are not included in our database. 4

5 Due to the lack of data it was impossible to simulate minimum pension support. Therefore we use data from the dataset. At this stage, a macro validation of the model and its database is considered, as evident in Table 1. Child is overestimated when calculated with the microsimulation model by somewhat more than 10%, which is evident from the table below. On the contrary, the average child from the microsimulation model and administrative sources almost coincide. The calculated state scholarship almost coincides with the aggregate scholarship data from administrative sources. By contrast, the calculated social assistance and data on minimum pension support 2 do not match the aggregates from administrative sources as well as the other two s, but the results are still satisfactory. Table 1: Macro validation of the main simulated s in a replication of the 2007 system in 1000 Administrative data Calculated data Index in 1000 in 1000 Child Social assistance * Minimum pension support State scholarship Notes: *The difference in the number of social assistances might be misleading as external statistics refer to the average monthly number of, while the number calculated with the microsimulation model refers to the total annual number of in Sources: Own calculations, microsimulation model; Ministry of Labour, Family and Social Affairs, 2011; Institute for Pension and Disability Insurance, Thirdly, all incomes are updated to 2010 and the s system in 2010 is replicated to introduce the changes to legislation between 2007 and Table 2 shows that the simulation results are less satisfying when 2010 is considered. The calculated amount of child and the number of are overestimated by more than 15%, though the average amount of child almost coincides with the average child from administrative records. Similarly, all other calculated aggregates of social s are overestimated, while only the calculated state scholarship coincides with the administrative data. The calculated minimum pension support 3 is very underestimated which can be explained by the underestimated number of in the sample in the first place and the changed legislation since The calculated social assistance is considerably overestimated in terms of aggregate numbers and in terms of the average level. The lion s share of this discrepancy lies in the already mentioned strong assumptions and data limitations which are the basis 2 Minimum pension support could not be simulated for the policy year 2007 as data on the pension qualifying period are not included in the sample but are essential for calculating the. Therefore, data taken directly from the sample are taken into account. 5

6 for the social assistance simulations (implying that incomes are received at the same rate and property data are missing). In addition, the income structure most probably changed between 2007 and 2010 as the minimum wage rose by 22.9% in March This increase mainly influenced people at the low end of the income distribution and the updating of incomes could not capture this structural change 4. Table 2: Macro validation of the main simulated s in a replication of the 2010 system in 1000 Administrative data Calculated data Index in 1000 in 1000 Child Social assistance Minimum pension support State scholarship Notes: *The difference in the number of social assistances might be misleading as external statistics refer to the average monthly number of, while the number calculated with the microsimulation model refers to the total annual number of in Sources: Own calculations, microsimulation model; Ministry of Labour, Family and Social Affairs, 2011; Institute for Pension and Disability Insurance, Finally, the new legislation concerning social s in 2010 is simulated. Then the percentage change between the calculated aggregate amounts considering the new legislation and the calculated aggregate amounts considering the current system in 2010 is estimated. The effects of changes foreseen at the aggregate level are estimated for 2010 by applying this percentage change to administrative aggregate data for Thus, data limitations and assumptions are mitigated to a certain extent. 3. REFORM CHANGES CONCERNING SOCIAL BENEFITS After the reform is implemented all social s 5 will be governed by the Exercise of Rights to Public Funds Act and the Financial Social Assistance Act. The new legislation changes family income which is the basis for the income test from gross family income to net family income, which also includes social s received. 3 It also could not be simulated for policy year The changed income structure does not influence the PIT simulations as much as social s. PIT calculated by the model coincides very well with the recorded PIT. The actual aggregate amount of PIT in 2004 was 3.9% higher than in our simulation 5 We only focus on the four most important s. However it should be noted that the microsimulation model was used to simulate other s/subsidies, namely: childcare subsidies, subsidised school meals, subsidised school transport, paid compulsory health insurance and entitlement to the full value? of health services. 6

7 Child is currently a means-tested, held by one of the parents for a child residing in Slovenia. The right to a child is retained until the child reaches 18 years of age, as well as for the period in which the child continues with full-time education, but only until the child reaches 26 years of age. Child is only paid in case the income per family member, in the calendar year prior to submission of a claim, was below the average wage in Slovenia. As already mentioned, a net income approach is taken into account under the new legislation, but the income brackets remains practically the same in absolute terms. The new legislation also reduces the age of a child s entitlement. The right to a child is held only until the child reaches 18 years. Besides, the child for eligible students included in higher secondary education (aged less than 18 years and with an income per family member below the average wage) is higher which is evident from Tables 3 and 4. Table 3: Monthly amounts of child in 2010 current legislation Monthly child for a child from birth to the end of primary Gross family income per family member school, in as a percentage of the average gross wage 3rd and every 1st child 2nd child subsequent child to 15% more than 15% to 25% more than 25% to 30% more than 30% to 35% more than 35% to 45% more than 45% to 55% more than 55% to 75% more than 75% to 99% Source: Ministry of Labour, Family and Social Affairs, Table 4: Monthly amounts of child in 2010 new legislation Net family income per family member as a percentage of the average net wage Monthly child for a child from birth to the end of primary school, in 3rd and every 1st child 2nd child subsequent child Monthly child for a child included in upper secondary education, in 3rd and every 1st child 2nd child subsequent child to 18% more than18% to 30% more than30% to 36% more than36% to 42% more than42% to 53% more than53% to 64% more than64% to 82% more than82% to 99% Source: The Exercise of Rights to Public Funds Act,

8 The increased amounts of child for children (under 18) included in upper secondary education are supposed to (at least partly) compensate for the abolition of the state scholarship for most upper secondary students. Currently, all upper secondary and tertiary students might be entitled to a state scholarship. On the contrary, only students (higher secondary and tertiary) aged 18 years or more might be eligible for a state scholarship under the new legislation. This means that the majority of upper secondary students will not receive any scholarship. Besides, the reform introduces a few more important changes: a) the family income used for the income test will be net family income with received child s included; b) family income will be compared to the average net wage instead of the minimum wage; c) the income threshold will be raised significantly; d) the monthly amount of the scholarship will be increased; and e) scholarship amounts will not differ according to education levels (upper secondary and tertiary students) as evident from Tables 5 and 6. Table 5: Monthly amounts of state scholarship current legislation Gross family income per family member as a percentage of the minimum wage Students in upper secondary education Basic scholarship Income supplement Students in tertiary education Basic scholarship Income supplement To 25% More than 25% to 40% More than 40% to 50% More than 50% to 60% More than 60% to 65% Source: Ministry of Labour, Family and Social Affairs, Table 6: Monthly amounts of state scholarship new legislation Net family income per family member as a percentage of the average net wage Basic scholarship with income supplement Students in upper secondary education Students in tertiary education To 30% More than 30% to 36% More than 36% to 42% More than 42% to 53% More than 53 to 64% Source: The Exercise of Rights to Public Funds Act, The Financial Social Assistance Act defines a new basis for ensuring social security for the most vulnerable population groups: a) family income which is the basis for the income test will include child s received and state scholarships; b) the lowest amount of financial social assistance will be increased; and c) an activity supplement will be granted to beneficiaries of financial social assistance working or taking part in activities or programmes aimed at employment. 8

9 Currently, the people entitled to cash social assistance do not have any income or receive income below the family minimum income. In the case of no income, the entitled persons receive the full stated amount; otherwise, they are entitled to receive the difference between their own income and the family minimum income. The minimum income for a family is obtained by multiplying the basic amount of minimum income (currently ) by the weighted number of family members. Table 7 shows the amounts of minimum income and weights in the current and forthcoming system. Table 7:Weights for the social assistance calculation according to the current and new legislation Current New legislation legislation Basic minimum income Weights: First adult in the family 1 1 Activity supplement for the first adult in the family working full time Activity supplement for the first adult working part time Adults aged 18 to 26 years living with their parents and registered as unemployed at the Employment Service of Slovenia Single persons permanently incapacitated for gainful employment or older than 63 years (women) or 65 years (men) living at the same address as other persons who are not their family members Adults in institutional care 0 1 All other adults in the family Activity supplement for all other adults in the family working full time Activity supplement for all other adults in the family working part time The oldest child in the family not included in upper secondary education All other children in the family not included in upper secondary education The oldest child in the family included in upper secondary education All other children in the family included in upper secondary education Sources: Ministry of Labour, Family and Social Affairs, The Financial Social Assistance Act, The same Act abolishes the state pension 6 and governs minimum pension support which will be completely transformed. Minimum pension support is currently granted to pensioners with low incomes and an incomplete contribution period, provided they fulfil a set of conditions, such as: a) their pension is lower than a certain threshold (81.6% of the lowest pension base, i.e in 2011); b) the family in which the pensioner lives does not have other income sources which would allow a minimum standard of living; and c) the family has negligible property. The amount depends on the contribution period, pension and number of family members (in the case of a survivor pension). Under the new legislation minimum pension support will be granted to all people permanently incapacitated for gainful employment or older than 63 years (women) or 65 years 6 The state pension will be replaced by social assistance and minimum pension support. 9

10 (men), irrespective of whether they are pensioners or not 7. The order of assessing the eligibility for different s will be changed. Currently, social assistance comes after minimum pension support whereas the new regulation will change the order of these two s. Eligibility for minimum pension support will be assessed similarly as for social assistance. The minimum income for eligible persons will be increased: 0.56 times the basic minimum income for a single person, first or only adult person in a family; and 0.28 times the basic minimum income for other adult persons in the family. 4. RESULTS The main results of the forthcoming reform are presented in Table 8, which includes comparisons of administrative aggregate data on main social s in 2010 and calculated aggregate data concerning these s on the assumption that the new legislation is enacted in Table 8: Calculated change in aggregates of simulated s Social s Administrative data, 2010 Calculated data, new legislation, 2010 Change in % Change in % Chid 374, , , , Change in % State scholarship (upper secondary 24,877 42, , , students) State scholarship (tertiary students) 14,609 30, , , Social assistance 48, , , , State pension 15,443 33, Minimum pension support 44,747 52, , , , , Sources: Own calculations, microsimulation model; Ministry of Labour, Family and Social Affairs, 2011; Institute for Pension and Disability Insurance, The results shown in Table 8 reveal that the reform of social s will lead to higher government expenditure as the aggregate amount of s will rise by 17.3%. We should keep all the assumptions and data imperfections in mind, which means that the predicted numbers only provide some orientation regarding future developments. The number of children entitled to a child will be reduced by 20% since children aged 18 years and more will lose the right to these s. However, the average amount of child will increase, mostly due to the higher amounts of 7 The name of the should be changed, for example to minimum elderly support. However, we keep the name to preserve the official naming. 10

11 child for upper secondary students. More than half of all upper secondary students will lose the right to a state scholarship, but those keeping this right will receive higher scholarship amounts. On the contrary, the state scholarship for tertiary students will grow substantially in terms of both the number of and average amounts. Similarly, social assistance and minimum pension support will also increase significantly. The rise in social assistance is expected due to a higher minimum income, higher weights and activity supplements, while the increased number of minimum pension support can be explained by widening the circle of beneficiaries. The three different measures Gini coefficient, squared coefficient of variation and the Atkinson index and squared coefficient of variation 8 presented in Table 9 reveal that the overall inequality will slightly decrease when the new legislation is enacted. However, inequality changes were not uniform across the population; inequality among children will increase while inequality among persons aged 63 years or more will decrease. Table 9: Income inequality measures based on household equivalent disposable income 9 Inequality measure Current legislation New legislation All Children 10 Elderly (63+) All Children Elderly (63+) Gini Atkinson ( 2) I Source: own calculations, microsimulation model. Additional results of the microsimulation are presented in Table 10. Individuals are aggregated into 10 decile groups regarding household equivalent disposable income. The first column presents annual average disposable income, while all the other columns present the shares of a particular social by income deciles. The results in Table 10 show that the reform leads to a relatively modest increase in equalised income and a change in the income distribution since incomes at the bottom of the income distribution will increase at an above-average rate. After the new legislation is enacted, more social assistance will be disbursed to persons in the second and third income deciles. Similarly, the state scholarship will be spread across the first seven income deciles, whereas it is currently focused on the first three income deciles. On the contrary, minimum income support is presently not disbursed to people in the first income decile as the right to this is linked to the right to a pension. The 8 All three measures are calculated according to Cowel (1977). 9 Household disposable income is calculated as the sum of net income subject to PIT and income not subject to PIT (social transfers). To calculate equivalent disposable income, the OECD s equivalence scale is used, which gives a weight of 1 for the first adult (over 16 years), a weight of 0.7 for the second and subsequent adult and a weight of 0.5 for any child. 10 A child is defined as a person below 18 or 26 years if they are a full-time student. 11

12 reform will change the nature of this which will focus largely on persons from the first two income deciles. Table 10: equalised income and shares of social s by income deciles under the current and new legislation Income decile equalised income, in Social assistance Child Income pension support Current legislation State scholarship (upper secondary students) State scholarship (tertiary students) 1 1, , , , , , , , , , , , New legislation 1 2, , , , , , , , , , , Source: own calculations, microsimulation model. To avoid a (too) big volume of information, in the rest of this section we only compare the income situation of different family types after the change in legislation. The results in Table 11 reveal that the reform will improve the income position of approximately one-third of families with pre-school children, and make it worse for the another third of families with children. Families with upper secondary students will experience a drop in the average amount of social s, although the share of families better off is higher than the share of families worse off. Conversely, the elderly will receive considerable higher social s on average, although the share of families with worse off elderly is relatively high. The reason lies in the changed nature of this and the redistribution of minimum pension support. 12

13 Table 11: effect of the forthcoming reform, by family type Families with Families better off, % Families worse off, % Annual amount of all s per equivalent adult, current legislation, in Annual amount of all s per equivalent adult, new legislation, in Pre-school children Children in primary school Students in upper secondary education ,149 1,069 Students in tertiary education ,050 Persons aged 63 or more ,010 1,381 Notes: Only families receiving social s under the current or new legislation are included. Source: Own calculations, microsimulation model. 5. CONCLUSIONS New legislation governing social transfers was adopted in July 2010 and is supposed to be enacted in June The Slovenian social transfers system has become characterised by claims that the system was regulated by too many different acts and institutions which led to unharmonised official records. Besides, work incentives are low and the accumulation of s is possible. The new legislation changes the eligibility rules for most social s which will influence the income position of beneficiaries and the income distribution. To a certain extent, the new legislation is based on estimations performed by the microsimulation model which was constructed for the purposes of the reform. Several administrative databases were merged to form a single database which contains a sample of 6% of the Slovenian population and represents a comprehensive fundament for the microsimulation. As the results in the paper show, the expected consequences of the reform are relatively extensive and give s mostly to individuals and households at the bottom of the income distribution. However, this reform will harm the government budget in the short run as expenditure on the simulated social s is supposed to rise by million (by 17% at current expenditure for the simulated s). At the moment Slovenia faces an unfavourable financial situation. The general government deficit in 2010 was estimated at 1,987 million or 5.5% of GDP and general government consolidated gross debt at the end of 2010 at 13,704 million or 38.0% of GDP (Statistical Office of the Republic of Slovenia, 2011). Therefore, the Slovenian government postponed the enforcement of both Acts to the beginning of An improperly established information system and inadequate official records contributed to this postponement. 13

14 REFERENCES Cowell F.A. (1977), Measuring Inequality, Oxford: Philip Allan. Čok M (2002) A Microsimulation Model for Slovenia s Personal Tax System, Economic and Business Review, 4(1), Institute for Pension and Disability Insurance (2008, 2010). Annual report (available at: Government of the Republic of Slovenia (2011). National Reform Programme Ljubljana, April Majcen B, Čok M, Verbič M and Kump N (2007) Razvoj in uporaba mikrosimulacijskega modela. Ljubljana: Inštitut za ekonomska raziskovanja. Majcen B, Bayar A, Mohora C, Čok M, et al. (2006) Analiza kompleksnih sektorskih in makro učinkov davčne reforme in reforme socialnih transferjev z uporabo dinamičnega modela splošnega ravnotežja slovenskega gospodarstva, Ljubljana: Inštitut za ekonomska raziskovanja. Ministry of Labour, Family and Social Affairs (2011). Statistics. (available at: Redmond G, Sutherland H and Wilson M (1998) The Arithmetic of Tax and Social Security Reform. A User s Guide to Microsimulation Method and Analysis, Cambridge: Cambridge University Press. Statistical Office of the Republic of Slovenia (2011). Main aggregates of the sector General Government, Slovenia, Ljubljana: Statistical Office of the Republic of Slovenia. (available at: Stropnik, N. and Stanovnik T. (2002). Combating Poverty and Social Exclusion - Volume 2, A Case Study of Slovenia. Budapest: International Labour Office. The Exercise of Rights to Public Funds Act. Official Gazette of the Republic of Slovenia, No. 62/201). The Financial Social Assistance Act. Official Gazette of the Republic of Slovenia, No. 61/