This chapter is divided in two parts. In the first part, it assesses and provides reform options for the design of the tax system including the design of the personal income tax, the basic tax allowance and social security contributions. An assessment of the sustainability of the social security system is also included. In the second part, the chapter assesses and provides reform options for the design of the benefit system including unemployment benefit, housing benefit, social assistance benefit and child benefits.
OECD Tax Policy Reviews: Lithuania 2022
OECD Tax Policy Reviews
3. The design of the tax and benefit system in Lithuania
Abstract
Introduction
Lithuania has a range of interacting policy levers that influence the tax and benefit system. The purpose of the policy lever map is to provide an overview of the various policy levers that are available to Lithuania to make adjustments to the tax and benefit system (Figure 3.1). The map highlights interactions within the system, for example that social assistance benefit is means-tested according to net income which is in turn reduced by the PIT and SSC rates. Similarly, the number of persons in a family can determine both the amount of social assistance benefit and housing benefit. The structure of this chapter covers each tax and benefit policy lever in turn.
Figure 3.1. A tax and benefit policy lever map for Lithuania
Note: Some other targeted benefits such as pensions and sickness and disability benefit are not included. Housing benefit here is shown only for rent. Compensations for heating costs, drinking water costs and hot water costs are not included (Būsto šildymo, geriamojo ir karšto vandens išlaidų kompensacijos), which covers a larger share of households-house owners. This benefit is outside the scope of OECD TaxBen model as the model considers only households-renters. Average wage refers to OECD secretariat calculations for 2020 (see Taxing Wages 2021). Figures are for 2021 unless otherwise stated. BSB refers to basic social benefit. MMS refers to minimum monthly wage. AMCN refers to amount of minimum consumption needs. SSI refers to state supported income.
Source: OECD analysis adapted from OECD Tax-Benefit Model for Lithuania 2020 and 2021.
Lithuania introduced a wide-ranging labour tax reform in 2019. The stated rationale for the reform was to make social insurance clearer, more attractive, to reduce the tax burden on labour and to make labour taxation the most competitive in the Baltic States (Government of the Republic of Lithuania, 2018). The reform included a number of changes to taxes and benefits including the following:
The introduction of a progressive PIT-rate system (with two PIT rate brackets) from a flat PIT-rate system.
Shifting SSCs from employers to employees.
A mechanical upward compensation adjustment in gross wages.
The introduction of employee and employer SSC ceilings.
An increase in the basic allowance and adjustment to the basic allowance withdrawal rate.
Increases in universal child benefit and additional child benefit.
Since the reform, a range of further tax and benefit policy changes were made in 2020 and 2021 including:
Increases in the top-PIT rate.
Cuts in the top-PIT rate threshold.
Cuts in the employee SSC ceiling.
Cuts in the employer SSC ceiling followed by abolition.
Part of the SSCs previously used to cover social insurance pensions was shifted to general taxation.
Further increases in, and a redesign of, the basic allowance.
Box 3.1. Labour tax reforms in Romania and Latvia
I. Romania undertook a tax reform that shifted almost all employer SSCs to the employee
In 2018, one year before Lithuania’s labour tax reform, Romania undertook a labour tax reform that similarly shifted almost all employer SSCs to the employee. Unlike Lithuania which reduced total SSC rates to 20.97% in 2019 (by cutting employer SSC rates by more than it increased employee SSC rates), Romania kept total SSC rates about the same at 37.25% in 2018. Romania increased employee SSC rates by 19pp to 35% in 2018 (up from 16% in 2017) while employer SSC rates were cut by a similar 20.5pp to 2.25% (down from 22.75% in 2017). Lithuania increased employee SSCs by 10.5p.p. to 19.5% in 2019 (up from 9.0% in 2018) but cut employer SSCs by a significantly greater 29.03p.p. to 1.47% in 2019 (down from 30.5% in 2018).
Romania’s reform increased tax revenues, unlike Lithuania’s which was broadly neutral. Employee SSC revenues in Romania rose markedly from 3% to 9% of GDP in 2018. This was partly offset by declines in employer SSCs and PIT, which fell to 1% (from 5%) and 2% (from 4%) of GDP respectively. Total labour taxes increased by 1% reaching 13% of GDP following the reform. In Lithuania, increases in employee SSC revenues (from 4% to 9%) and PIT revenues (from 4% to 7%) were offset by reduced employer SSCs (from 9% to 1%).
Romania employed a different wage strategy to Lithuania, but gross wages still rose. Lithuania increased gross wages mechanically in January 2019 so that net disposable income for most workers remained unchanged. Statistics Lithuania data show that average monthly earnings by 30% in Q1 2019 compared to Q4 2018, suggesting that Lithuanian firms compensated workers with even higher gross wages than required by the mechanical adjustment. Average monthly earnings have continued to grow in Lithuania since then up to the latest data in 2022, albeit more slowly. In Romania, while no such mechanism was employed and instead wage bargaining occurred between employees and employers, gross wages increased steadily in the years after 2018 and up to 20221. Gross wages likely increased in Romania as the shift in SSCs from employer to employee allowed firms to increase gross wages without further increasing total compensation of employees and, at the same time, workers demanded higher gross wages to compensate for higher employee SSCs (European Commission, 2019[1]) (For a discussion of employer and employee SSC incidence see Box 3.2).
Figure 3.2. Taxes as a % of GDP in Romania and Lithuania, 2016 - 2019
Note: PIT refers to taxes on individual or household income including holding gains. Employee SSCs refer to compulsory social contributions by households. Employer SSCs refer to social contributions by employers. Employee SSCs in Romania are comprised of 25% general health SSC and 10% health SSC in 2018 (which were previously 10.5% and 5.5% respectively in 2017). An employee SSC for unemployment was also abolished in 2017.
Source: Data on Taxation, European Commission.
II. Latvia introduced a tax reform that included transitioning to a progressive PIT system
Latvia introduced a labour tax reform in 2018. The reform had several aims including strengthening tax wedge progressivity and reducing the tax burden on low-income earners with no dependents.
As part of the reform, Latvia transitioned to a progressive PIT rate system. A flat 23% rate was replaced with three PIT rate brackets of 20%, 23% and 31.4%. To support the financing of healthcare services, the SSC rate increased by 1 p.p. from 34.09% to 35.09%, shared equally between employers and employees (employer SSC rate increased 0.5 p.p. and employee SSC rate increased 0.5 p.p.).
Following the reform, tax revenues remained broadly similar and the impact on progressivity was limited. After the reform, SSCs-to-GDP increased from 8.4% to 9.2% while PIT-to-GDP fell from 6.6% to 6.0%. The decline in PIT-to-GDP was partly due to a relatively large proportion of taxpayers in the lower tax bracket that pay PIT at the lower rate. Overall, PIT and SSC revenues as a share of GDP remained broadly unchanged at 15.1% in 2018 (in 2019, PIT and SSCs both increased to 9.6% and 6.5% as a share of GDP respectively). The directionality of these PIT and SSCs changes had been anticipated by the authority’s pre-reform. Some evidence has pointed to a relatively limited impact on inequality following the reform as the progressive PIT schedule tended to benefit middle-income earners and was less well targeted at poorer households (Ivaškaitė-Tamošiūnė et al., 2018[2]).
1. According to data from Statistics Romania, monthly gross average earnings increased from 4143 Romanian leu in January 2018 to 6031 Romanian leu in January 2022.
The design of the tax system
Lithuania operates a worldwide personal income tax system. Lithuanian residents are taxed on their worldwide income while non-residents are taxed on the basis of the territoriality principle and, as such, are only taxed on income from Lithuanian sources. Worldwide taxation for personal income is more prevalent around the world than territorial taxation (Shum, Fay and Lui, 2017[3]).
A new rule in 2022 requires employers to pay all employment-related payments into employee bank accounts. Lithuania is a relatively high cash economy according to survey data from the European Central Bank. This policy brings Lithuania in line with many countries that have introduced stricter enforcement making it compulsory for employers to pay salaries through a bank account.
SSCs are not deductible from the PIT base, which is atypical among OECD countries. A common form of interaction between PIT and SSCs in OECD countries is the full or partial deductibility of SSC payments from the PIT base. For example, in Estonia, employee SSCs for unemployment insurance are deductible from PIT as a standard tax relief. In Poland, an allowance is provided for all social insurance contributions paid by the taxpayer and a tax credit is available for health insurance contributions. In Germany, SSCs are deductible up to specific ceilings. In Lithuania however, SSCs are not deductible from the PIT base.
Non-deductible SSCs imply lower disposable incomes. As SSCs are deductible from the PIT base in most OECD countries, SSC rate hikes narrow the PIT base and mechanically reduce effective PIT rates. As SSCs are not deductible in Lithuania, workers face relatively higher marginal and average PIT rates and have less disposable income relative to allowable SSC deductibility. The non-deductibility of SSCs in Lithuania support the case for not taxing future SSC pension benefits under the PIT as they have effectively already been taxed through higher PIT. Similarly, while shifting from employer to employee SSCs would mechanically reduce PIT revenues to some degree in most OECD countries due to lower taxable personal income (given constant total labour costs for the employer), it would not do so in Lithuania.
Lithuania transitioned to a progressive personal income tax system. Lithuania transitioned from its flat 15% PIT rate system to a progressive PIT rate system in 2019 with a lower 20% rate and an upper 27% rate. In 2020, the top-PIT rate was increased to 32% and the lower PIT rate remained the same (Table 3.1, Table 3.2).
Table 3.1. The PIT rate schedule
Note: In 2020, a 20% PIT rate applies to taxable income up to EUR 104 278 per year (EUR 8 698 per month) or 84 average wages (AW)and the PIT rate for the part exceeding this limit is 32%.
Source: OECD analysis.
Average and marginal PIT rates are relatively compressed. At lower incomes (67% of AW), average PIT rates in Lithuania rank among the top 10 OECD countries (Figure 3.3). Relatively high PIT rates at low incomes contribute to the work disincentive in Lithuania (see chapter 4). At higher incomes (AW and 133% of AW), average PIT rates are in the middle of OECD countries. Average PIT rates are relatively compressed ranging from 14% to 20%, indicating relatively little PIT progressivity. Marginal PIT rates in Lithuania are in the middle of OECD countries. Marginal PIT rates are the same for the income levels shown, again reflecting not much progressivity.
Figure 3.3. PIT rates are relatively compressed
Marginal and average PIT rates in OECD countries, 2021
Note: Analysis relates to an average single individual without children. The marginal rates are expressed as a percentage of gross wage earnings.
Source: OECD tax database.
Average PIT rates are quite high at incomes below the average wage. Average PIT rates for single individuals in Lithuania are above the OECD average at lower incomes (below the AW) but below the OECD average at higher incomes (167% of AW) (Figure 3.4 Panel A). Single and couple Lithuanian parents with two children have markedly higher average PIT rates than the OECD average. The average PIT rate in Lithuania is above the OECD average and several peer countries between about 1/3 of the AW and the AW (Figure 3.4 Panel B). Above the AW, the average PIT rate in Lithuania remains flat while the OECD average PIT rate rises.
Figure 3.4. Average PIT rates are quite high at low incomes
Average PIT rates, 2021
Note: In panel A, ch = children. 1 = two-earner couple. In panel B, data refers to a single individual with no children. In panel B, OECD average refers to an unweighted average of the average PIT rate by percentile in 29 OECD countries for which data are available.
Source: OECD tax-benefit model. Model version 2.4.0. Taxing Wages 2022.
Microdata analysis confirm that many lower income employees face relatively high tax burdens. Over 1/3 of employees earn below half the AW (Figure 3.5, Panel A). The average backward-looking tax wedge by percentile calculated using the microdata (i.e. tax wedge 2 in Figure 3.5, Panel A) and the hypothetical tax wedge (i.e. tax wedge 1 in Figure 3.5, Panel A) highlight a relatively flat tax burden across the employment income distribution and high tax burdens at low incomes (due to the employer SSC floor) (Figure 3.5, Panel A). In addition, mean PIT and SSCs paid by employment income in the microdata are consistent with the tax rules (Figure 3.5, Panel B) - the PIT share rises over time because the PIT rate rises steadily as the BA is tapered and the SSC shares are flat due to the flat employee and employer SSC rates.1
Figure 3.5. Many lower income employees face high tax burdens
Tax wedges, share of taxpayers and taxes paid, by total income as % of AW
Note: Tax wedge (1) is based on the OECD Taxing Wages model in 2021 for a single employee using the same AW. Tax wedge (2) is the mean backward-looking tax wedge by percentile calculated using the microdata in 2019. The total number of taxpayers with positive employment income is 46 389 in the sample microdata (i.e. employees). The microdata is for the year 2019. Some key tax rules have changed since 2019. In 2019, the BA was EUR 3 600 for annual incomes up to 12 MMS (EUR 6 660). The lower PIT rate was 20% (up to 120 AW i.e. EUR 136 334) and the upper PIT rate was 27%. In 2021, the BA had increased to EUR 4 800 up to 12 MMS (i.e. EUR 7 704). The lower PIT rate is 20% (but up to 60 AW i.e. EUR 81 162) and the upper PIT rate increased to 32%. The employee and employer SSC rates of 19.5% and 1.47% remained the same. In 2021, the employee SSC ceiling was cut to 60 AW and the employer SSC ceiling was abolished.
Source: OECD analysis of microdata 2019.
Lithuania has halved the top-PIT rate threshold in recent years but it remains high. Lithuania has cut the top-PIT rate threshold (i.e. the income level at which the top-PIT rate starts to apply) from about 10 to 5 annual AWs between 2019 and 2021 (Table 3.2).2 The top-PIT rate threshold as a share of the gross wage in Lithuania has gone from being one of the highest among OECD countries to converging towards the average in OECD countries (Figure 3.6). However, the top-PIT rate threshold remains high at EUR 8 698 in 2020 relative to the monthly average wage of EUR 1 400 (OECD estimate3) and the monthly minimum wage (MMW) of EUR 1 400. The cuts to the top-PIT rate threshold between 2019 and 2021 (which increase the tax burden on higher earners) should be considered in the context of employee and employer SSC ceilings, which were jointly reduced to the same AW multiples during that period (which reduce the SSC burden on higher earners, see Table 3.8 in the SSC section).
Table 3.2. Lithuania has cut top-PIT rate thresholds and increased top PIT rates
PIT rates and higher PIT-rate thresholds in Lithuania, 2018 - 2021
Notes: Monthly AW are converted to annual AW by dividing by 12. In 2021, 60 * EUR 1 352.70 = EUR 81,162 per year. In 2020, 84 * EUR 1241.4 = EUR 104 277.60 per year. In 2019, 120 * EUR 1,136.2 = EUR 136,344.
Source: OECD analysis.
Figure 3.6. The top-PIT rate threshold has converged towards the OECD average
Top PIT thresholds as a share of gross wages in Lithuania and selected OECD countries, 2020
Note: Gross wages incomes are based on OECD secretariat calculations from Taxing Wages 2021 publication. In Lithuania, the average wage is EUR 12 095 in 2019 and EUR 16 426 in 2020. Average wage in Lithuania in 2021 is assumed to be the same as 2020. *OECD refers to an weighted mean average for the selected countries shown.
Source: OECD tax database; Taxing Wages 2021.
Developments in labour tax revenues
In 2019, the year of the labour tax reform, labour tax revenues remain broadly unchanged. PIT and total SSC revenues as a share of GDP increased only modestly in 2019 (Figure 3.7). This occurred because the PIT-to-GDP ratio and the employee-SSCs-to-GDP ratio rose from 4.1% to 7.2% and 2.5% to 7.3% respectively, but they were mostly offset by the employer-SSCs-to-GDP ratio decline from 8.5% to 0.7%. After that in 2020, employee-SSCs-to-GDP increased while employer-SSCs-to-GDP and PIT-to-GDP remaining broadly unchanged.
The reform involved the most significant reshuffling of the labour tax mix in decades. Employee SSC revenues as a share of GDP have gone from playing a minor role in 1995 to a major role in 2020 and vice versa for employer SSC revenues (Figure 3.7). To explore the consequences of this SSC reshuffling, Box 3.2 considers the role of the tax incidence concept for employee and employer SSCs. Box 3.1 examines the case study of Romania in 2018 which also shifted most of its SSCs from the employer to the employee.
Figure 3.7. Labour taxes were broadly unchanged in 2019
Labour tax revenues as % of GDP, 1995 - 2020
Note: PIT relates to incomes and profits of individuals (code 1110). Employee and employer SSCs relate to OECD revenue statistics codes 2100 and 2200 respectively.
Source: Revenue Statistics 2021 (2021[4]).
The personal income tax system could be more progressive at the top
The effective tax rate is high in international comparison but low when employer SSCs are included. Average effective tax rates (AETRs) for a single individual with no children in Lithuania in 2021 are the third highest in the OECD at low and middle incomes (50% and 100% of the AW respectively) (Figure 3.8 Panel A). AETRs at low incomes are driven mostly by employee SSCs (see section 3.2). When employer SSCs are included in the ETRs (which is not a typical measure used in OECD Taxing Wages that brings the calculation closer to a tax wedge type measure), Lithuania’s AETRs do not change by much (Figure 3.8 Panel B). Lithuania’s AETRs are now lower when compared to other OECD countries because the AETRs of other countries have increased by more (due to Lithuania’s relatively lower employer SSCs). A similar trend occurs at high incomes. At 200% of AW, measured without employer SSCs, AETRs in Lithuania are the 8th highest in the OECD countries shown in 2021. A single individual with no children at 200% of the AW in Lithuania faces a lower AETR than in Germany but higher than in Estonia, Latvia and Poland (Figure 3.9). However, when employer SSCs are included, a single individual earning 200% of the AW in Lithuania is then ranked the 4th lowest among the OECD countries shown in 2021.
Figure 3.8. Effective tax rates in international comparison
Effective tax rates including and excluding employer SSCs, single individual with no children earning 50% and 100% of the AW, 2021
Source: Taxing Wages 2022 tax decomposition data. OECD tax benefit model 2021.
AETRs in Lithuania are relatively flat at higher incomes. AETRs in Lithuania are compressed, ranging from 30% at low incomes (50% of the AW) to 40% at high incomes (250% of the AW) (Figure 3.9). AETRs at high incomes (above the AW) are also quite flat. In these two respects, Lithuania’s ETR structure is not dissimilar to Germany. In addition, the PIT rate on low-incomes (67% of AW) as a share of the PIT rate on high-incomes (167% of AW) is the fifth-highest in the OECD at 70% in 2021, indicating relatively little PIT progressivity. Including employer SSCs in AETRs makes Lithuania’s ETRs comparatively lower but the ETR structure remains quite compressed and flat.
Figure 3.9. Effective tax rates are not very progressive at the top
Effective tax rates, single individual with no children, gross income as % of AW, 2021
Note: The effective tax rate is the share of gross income that is due to the government as income tax and employee social-security contributions, minus social benefits received. The effective tax rate including employer SSCs is calculated by adding employer SSCs as a share of gross wage earnings to the effective tax rate.
Source: Taxing Wages 2022 tax decomposition data. OECD tax benefit model 2021.
Rapidly rising wages may induce bracket creep. Rapidly rising inflation and nominal wage growth in Lithuania (Figure 2.7) means that taxpayers will reach the higher marginal tax rate more quickly than previously (i.e. bracket creep). Consequently, the non-indexation of PIT brackets (i.e. PIT brackets are not systemically increased to account for wage growth or inflation) effectively increases the tax burden.
The introduction of a middle PIT rate bracket could be considered
Few employees face the top PIT rate due to the high top PIT rate threshold. Despite the top PIT rate threshold being cut from 10 to 5 annual AW between 2019 and 2021 (i.e. to EUR 81 162),4 it remains high in the context of the employment income distribution. To be liable for the top PIT rate, employees would need to earn in the top 1% of employment income as measured by employment income percentiles (i.e. the average employment income in the top 1% is EUR 7 542 per month or EUR 90 501 per year) (Figure 2.7, Panel C). The top PIT rate threshold is also high in a total income distribution context as it is similar to the mean total income of the top 2% of employees as measured by total income percentiles (i.e. EUR 83 122, see (Figure 2.8, Panel B).
A new middle PIT rate bracket could be considered. The introduction of a third middle PIT bracket could be considered which would effectively act as a de facto top PIT bracket if the current top PIT rate bracket were retained (given that so few employees face the current top PIT rate). Some combination of a reduced top PIT rate threshold and the introduction of a third middle PIT bracket could also be envisaged. A PIT rate threshold targeted at the top 10% and 5% of employees (rather than the current top PIT rate which targets only the top 1%) would correspond to a top PIT rate threshold of about 1.5 and 2.0 times the annual AW (instead of the current 5 times the AW) (Figure 2.8, Panel C).
Introducing a middle PIT rate bracket would raise PIT revenues. The introduction of a middle PIT bracket at 2 annual AW at PIT rates of 24%, 26% and 28% would raise PIT revenues by 1.5%, 2.3% and 3.2% respectively (Figure 3.10, Panel A). Almost all of the additional PIT would be raised from employees in the top income quintile and particularly in the top decile (Figure 2.8, Panel B). The introduction of a middle PIT bracket would lead to a modest reduction in income inequality as measured by the S80/S20 ratio (Figure 3.10, Panel C). Setting the middle PIT rate threshold at 1.5 annual AW could also be considered. The simulations in Figure 3.10 are based on EUROMOD, which is a tax-benefit microsimulation model for the European Union that is maintained, developed and managed by the Joint Research Centre (JRC) of the European Commission (see the note in Figure 3.10 for details).
Cutting the top PIT rate and simultaneously aligning the SSC ceiling with the top PIT rate threshold would reduce overall PIT and SSC revenues. Cutting the top PIT rate threshold from 5 to 4 annual AW and simultaneously aligning the SSC ceiling with the PIT rate threshold would raise PIT revenues but reduce SSC revenues. Overall PIT and SSC revenues would decrease modestly and income inequality would increase modestly (the S80/S20 ratio would increase by 0.0023 percentage points from 10.7604 to 10.7627).
Figure 3.10. The introduction of a new middle PIT rate bracket would raise PIT revenues and reduce inequality
Simulating the introduction of a middle PIT bracket, 2 annual average wages
Note: EUROMOD is a tax-benefit microsimulation model for the European Union that is maintained, developed and managed by the Joint Research Centre (JRC) of the European Commission, in collaboration with Eurostat and national teams from the EU countries. This report makes use the beta-version of the model for Lithuania, based on version I3.0+, adapted to run on the input data derived from national administrative registers. This version is not available to the public. The tax-benefit system simulated in the baseline scenario refers to the one in force in 2022. The underline input data come from national administrative registers for year 2019. Uprating factors are used to bring the income values from the income reference period to the policy year, in this case from 2019 to 2022. Simulations provide static results, as no labour supply or general equilibrium effects are estimated in EUROMOD.
Source: Simulations performed by the EUROMOD team of the European Commission’s Joint Research Centre (JRC).
Box 3.2. Employer and employee SSC incidence
The theoretical literature argues that formal tax incidence (who legally is obliged to pay tax) is irrelevant for effective tax incidence (who eventually bears the burden of the tax), at least in the long run. This line of argument suggests that the same given increase in employee SSCs or employee SSCs is immaterial of who the tax burden falls on. The effective tax incidence (i.e. who eventually pays the tax or is ‘out-of-pocket’, irrespective of who has paid the cash) of an SSC increase is ultimately borne by the employee, typically through the mechanism of reduced wages, and this occurs regardless of whether the tax is paid by employers (through an employer SSC increase) or employees (through an employee SSC increase). In standard economic theory, this occurs because employees are more inelastic than employers to SSC changes. In this context, inelasticity means that workers have few good alternatives to work because they have to ‘pay the bills’ somehow and are unlikely to have sufficient capital to live off that exclusively. By contrast, employers have more good substitutes – they can swap labour for capital or relocate to where labour is cheaper. For example, in a meta-analysis, Melguizo and Gonzalez-Paramo (2013) find that two-thirds of the incidence of labor taxes generally fall on employees, albeit there is a very wide range of estimates.
The empirical literature finds that, in general, SSC cuts lead to higher employee wages but also that the tax incidence may vary across different settings (e.g. short run versus longer run, SSC rate increases versus decreases, changes in employee versus employer SSCs, etc.). While relatively few studies directly examine the incidence of employer and employee SSCs, some of the more recent and empirically rigorous studies indicate that there is relatively limited shifting of SSCs to employees in the form of changed wages. A number of studies use macroeconomic evidence based on cross-country regressions (such as (Arpaia and Carone, 2004[5]) find evidence that the incidence differs in the short-term (perhaps reflecting short-term stickiness of nominal wages). There are also a number of quasi-experimental studies. (Gruber, 1995[6]) finds evidence that a decrease in pension SSCs in Chile led to an equivalent increase in wages that is consistent with the full incidence of SSCs on employees. (Anderson and Meyer, 1997[7])find that companies with larger employer SSC reductions increase employee earnings roughly one-for-one (and employment is unaffected). They interpret this as evidence that the incidence of employer SSC is largely or fully on workers. There finding may also, at least in part, be explained by behaviour shifts (i.e. employees working harder due to a SSC cut, which induces employers to increase salaries). For example, using three large increases in SSCs in France between 1976 and 2010, (Bozio, Breda and Grenet, 2017[8])find very limited shifting of SSCs to employees in the form of lower wages. They interpret this as wage stickiness prevents employer SSCs from being shifted to employees in the short-term. (Adam, Phillips and Roantree, 2019[9])find that statutory incidence matters for the short-term economic effects of SSCs in the UK. They interpret their findings as evidence that employees change their hours in response to SSCs, but that in the short to medium term at least, the formal incidence of SSCs can matter for their behavioural impacts and economic incidence. (Bozio, Breda and Grenet, 2019[10])find that the tax incidence falls on employers for reforms that have no tax-benefit linkage whereas the incidence of SSCs falls on employees in reforms with strong tax-benefit linkages. A tax-benefit linkage distinguishes SSCs from other labour income tax. If workers incorporate in their labour supply decisions not only net wages but also expected benefits, behavioural responses may be reduced (because workers have even less incentive to shift to alternatives). In that case, SSCs with strong tax-benefit linkages are expected to be fully shifted to workers.
The design of the basic allowance
Lithuania increased the basic allowance in recent years. A personal basic allowance (BA) (i.e. a tax-exempt amount) is applied to employment income in Lithuania to calculate taxable income. The BA was EUR 350 per month in January 2020, the max BA was available up to EUR 607 per month and a tapering equation was applied thereafter (Table 3.3). In response to the COVID-19 pandemic, the BA was increased to EUR 400 in July 2020 and the withdrawal rate was modestly increased. This change was applicable for the whole of 2020. In 2021, the BA threshold was increased to EUR 642 and the withdrawal rate was modestly decreased. In 2022, the BA was increased to EUR 460 and the number of ‘break points’ was increased from one to two. Between 2019 and 2020, the rate of increase in the BA in Lithuania of 13% was higher than the rate of increase in some other OECD countries (including Canada +5.0%, Germany +3.6% and United Kingdom +0.6%).
The basic allowance as a share of the average wage is not dissimilar to that found in some other OECD countries. The maximum BA increased in Lithuania represented 27% of the AW in 2022. In Slovenia, Latvia, the Slovak Republic and Estonia, the BA represented 17%, 28%, 33% and 36% of the AW respectively in 2020.
Table 3.3. Lithuania has increased the basic allowance in recent years
Notes: 2020 refers to the basic allowance introduced in July 2002. ‘inc’ refers to individual gross income. If the BA as calculated by the formula is negative, it is assumed equal to zero. In 2022, there are two thresholds with two different associated BA formulas. AW for 2020 Jan, 2020 July, 2021 and 2022 are based Statistics Lithuania average gross monthly earnings for 2020Q1, 2020Q2, 2021Q1 and 2021Q4 2021KQ (latest available data), which are EUR 1 381, EUR 1 398.5, EUR 1 517.4 and EUR 1 679.3 respectively. MMW refers to minimum monthly wage.
Source: OECD analysis.
Lithuania operates a basic tax allowance withdrawal system, which is common in OECD countries. The withdrawal rate determines both the rate at which the BA is decreased for higher incomes above the maximum BA threshold and the income level at which the BA is no longer received at all. As incomes start to increase above the BA threshold, the amount withdrawn increases. Withdrawal-type systems are operated in different forms in several OECD countries including Slovenia, the Slovak Republic and Latvia.
The BA withdrawal rate can be used to increase the progressivity in the middle part of the income distribution. Since the withdrawal rate starts at the max BA threshold and ends when the BA is fully withdrawn, the withdrawal rate targets tax burden adjustments in the low to middle part of the income distribution. A withdrawal rate of zero implies no withdrawal and the BA becomes a flat amount that all taxpayers benefit from irrespective of their income. A withdrawal rate close to one implies a rapid rate of BA reduction so that effectively only low incomes can benefit from it. The increase in the withdrawal rate from 17% in 2020 to 26% in 2022 contributed to increased PIT progressivity.
The 2022 basic allowance redesign introduces more progressivity in to the tax system through both a higher basic allowance and a faster tapering rate as incomes rise. The BA 2022 reduces the PIT burden at low incomes and supports progressivity, reflected in the downward shifted PIT curve (Figure 3.11). An employee at 50% of the AW has seen their PIT fall from 11% to 9% to 7% in 2020, 2021 and 2022 respectively (Figure 3.11). Compared to the 2021 BA design, the 2022 BA design has a first ‘break point’ with a steeper slope and a second ‘break point’ that matches the 2021 BA design at higher incomes. The first ‘break point’ tapers more quickly than the 2021 taper rate as incomes are increased beyond the BA max threshold. The second ‘break point’ occurs at 118% of the AW and slows the rate of the BA taper to match that of the 2021 design.
Limitations of the basic allowance
The basic allowance is somewhat untargeted. While the BA is and has been an overall effective way to produce PIT progressivity in Lithuania, it is somewhat untargeted. Compared to the 2020 BA, the 2022 BA is withdrawn at a higher income, which may benefit some higher earners. As a result of the BA increase between 2020 and 2022, a high earner (150% of the AW) saw a modest average PIT rate cut from 19.1% to 18.7%. The BA is fully withdrawn at 189% of the AW in 2020 and at 203% of the AW in 2021 and 2022, after which point the average PIT rate converges to the 20% marginal PIT rate (and at higher incomes to the 32% PIT rate).
Most of the basic allowance benefit went to workers earning between the minimum and average wage. The large share of Lithuanian workers with wages between the MMW and the AW (Figure 2.9) led to this cohort receiving the largest share of the BA benefit. 57% of the BA in 2020 went to individuals earning between the MMW and the AW despite that the cohort represented only 44% of individuals. 29% went to those below the MMW (Figure 2.9). Therefore, the structure of the wage distribution contributed to most of the BA benefit going to middle rather than lower wage earners.
Figure 3.11. The 2022 basic allowance has a higher amount and a faster tapering rate
The amount of the basic allowance (EUR) and the percentage PIT rate in Lithuania, 2020 – 2022
Note: A caveat to the analysis show is that the BA is examined in each year in nominal terms using the average wage in 2020 (on the x-axis) but the real BA value is declining given inflation and rising wages in Lithuania
Source: OECD analysis.
The basic allowance design in 2020 and 2021 produce some marginal PIT rate spikes at low incomes close to the minimum monthly wage. The marginal PIT rate patterns in 2020 and 2021 are broadly similar (Figure 3.12, Panel A). The first spike in the marginal PIT rate occurs as employees first start to pay PIT (at 31% of the AW in 2021, slightly above the level of 2020 due the greater BA). The second spike occurs as the max BA starts to be tapered out (at about the MMW in 2020). The marginal tax wedge is higher than the marginal PIT rate but follows a similar pattern (Figure 3.12, Panel B).
The 2022 basic allowance design produces even higher marginal PIT rates spikes than in previous years. For the 2022 BA design, the first spike in the marginal PIT rate is the same magnitude as in 2020 and 2021 but it occurs at a higher income due to the higher BA amount. The second spike occurs after the max BA threshold and is larger in magnitude than in previous years due to the more rapid BA tapering.
The basic allowance design may produce modest poverty traps but for many workers. For both the marginal PIT rate and the marginal tax wedge between 2020 and 2022, spikes in the marginal tax rates occur at about 1/3 and 1/2 of the AW. A large share of Lithuanian employees earn wages at these incomes. These spikes could reduce the incentives for low-income employees to earn more income, to work more hours or to work harder beyond these thresholds (to avoid additional tax burdens). Modestly high marginal tax rates could also make working extra hours in the informal economy more attractive, which is not low or decreasing in Lithuania (Figure 2.4). The extent to which employees ‘bunch’ below these marginal tax rates should be examined further in the tax record data.
Figure 3.12. The basic allowance produces marginal tax rate spikes at low income levels
Marginal PIT rate and marginal tax wedge, based on basic allowance in 2020, 2021 and 2022
Note: The basic allowance amounts in EUR are set at 350, 400 and 460. The basic allowance thresholds in EUR are set at 607, 642 and 730 (and 1,678). The AW is set at the 2020 level of EUR 16,844 as calculated by the OECD secretariat and the MMW in 2020 is 607.
Source: OECD modelling.
Rising wages reduce the value of the basic allowance. Rising wages driven by inflation undermine the value of the BA. As wages rise, the max BA threshold is reached at a lower income of 44% of AW instead of 51%, implying higher effective PIT burdens on employees (Figure 3.13). Similarly at 50% and 100% of the AW, higher wages imply higher PIT rates of 9% vs 7% and 17% vs 16% respectively.
The basic allowance should be inflation-linked. The BA is not formally indexed to inflation. It is however set annually by the government informed by the increase in the MMW (the MMW is in turn set by the government, unions and employers indexed based on the MMW as a share of the forecast AW). Indexing the BA to inflation is relatively common internationally. A rationale for a BA generally is to help cover job-related expenses. Since job-related expenses rise with inflation, it follows that the BA should also be inflation-linked. Inflation in Lithuania was reported as among the highest in the OECD. The CPI reached 19% in May 2022. Linking the BA to inflation in Lithuania would maintain its value in real terms including for low-income workers.
Whether the basic allowance should be linked to wage growth is a policy choice, but it could come with PIT revenue risks. Linking the BA to a measure that might outgrow inflation, such as the MMW or the AW, is a policy choice. The purpose of a BA and a progressive PIT system is partly that a growing economy with rising wages will translate to higher PIT revenues. However, by linking the BA to increases in wages instead of inflation, a relatively larger share of workers could be exempt from paying any PIT when wages outpace inflation. It may be preferable from both a compliance and PIT revenue perspective that low-income workers pay a small PIT amount rather than none at all.
Figure 3.13. Rising wages reduce the relative value of the basic allowance and increases average effective PIT rates
The basic allowance and the average PIT rate, based on a simulated AW in 2020 and 2022
Note: 2020 average wage is based on the OECD secretariat calculations. This is increased by 9.9% in 2021 and a further 5.3% in 2022 based on Statistics Lithuania average wage increases in 2021. Note that in 2022 wages increased even more quickly.
Source: OECD analysis.
The design of the basic allowance may be contributing to income bunching
There is evidence of employee income bunching at the minimum monthly wage, particularly among older, married and female employees. The MMW is the most commonly reported employment wage income among employees (i.e. EUR 555 in 2019). 4.7% of employees report an income between 39% and 41% of AW (Figure 3.14, Panel A). Income bunching is higher for different employee types. In the aforementioned 39% to 41% of AW income range, income bunching is higher for married couples vs single individuals (5.0% vs 3.8%), for older employees aged over 50 vs employees aged under 35 (5.7% vs 3.5%) and for female vs male employees (5.6% vs 3.8%) (Figure 3.14, Panels B, C and D). Income bunching is particularly high for employees that are aged above 50, married and full-time employees (6.4%). Single and younger workers under the age of 35 are more likely to report income at very lower incomes between 0 – 5 of the AW ((Figure 3.14, Panels B and C).
Income bunching at the minimum wage coincides with a jump in the METR due to the loss of the maximum basic allowance as incomes rise. While much of the income bunching is likely due to the MMW itself, the sharp distributional spike in the share of employees reporting an income at the MMW occurs just before a spike in the METR, which is partly driven by the tapering of the max BA.5 Therefore, employees may be responding to this jump in the METR by reporting incomes below it at the MMW (i.e. ‘income bunching’) partly to benefit from the max BA and so as not to lose the BA at higher incomes. If this were the case, the greater degree of income bunching among older employees, married couples and women could suggest that these groups are particularly responsive to increased METRs.
Figure 3.14. There is evidence of employee income bunching at the minimum monthly wage
Bunching of employees at the MMW, % of employees, by employment income as % of AW
Note: Analysis based on 43,389 individual taxpayers with positive employment income. Of these, 55% are married and 29% are single. 33% are younger (i.e. 35 years of age or younger) and 34% are older (i.e. 50 years or older). 50% are female and 50% are male.
Source: OECD analysis of microdata 2019.
Medium-term PIT reform options
The introduction of an in-work benefit could be investigated further as part of future work
An in-work benefit could be envisaged. In-work benefits are welfare schemes designed to provide income supplement to needy families or individuals on the condition that they work (OECD, 2005[11]). In-work benefits (IWB) typically involve tax reductions or cash transfers that are conditional on labour market participation. The value IWBs, such as an earned income tax credit (EITC), commonly depend on the recipient’s earned income. The value of the credit increases gradually with income (phase-in) until it reaches a maximum credit amount (plateau). Beyond a certain earned income threshold (the phase-out threshold), the value of the credit is gradually reduced to zero. IWBs are typically means-tested and may be targeted at specific groups. When designed correctly, the IWBs have the potential to alleviate in-work poverty and increase work incentives and labour market participation for lower income workers (OECD, 2011[12]). Table 3.4 outlines selected advantages and limitations of introducing an IWB in Lithuania.
In-work benefits tend to help the unemployed find jobs and the unemployed face high poverty rates. High and rising poverty rates among the unemployed (Figure 2.11) suggest that disincentives to enter work or inability to work may be a more pressing policy concern for Lithuania than disincentives to progress in work. Most evidence suggests that IWB have positive employment effects when moving from unemployment to work (Chetty, Friedman and Saez, 2013[13]) (Blundell and Shephard, 2012[14]). This occurs because the substitution effect (workers can gain more per hour worked with an IWB) tends to outweigh the income effect (workers reduce hours as IWB raise disposable income). With regard to progressing in work, the evidence is more mixed and the effects tend to be smaller (OECD, 2011[12]). However, this may be less of a concern in Lithuania given high employment and participation rates and relatively lower poverty among the employed (Figure 2.3).
An in-work benefit could compensate for relatively high PIT and SSCs faced by low income workers. A refundable in-work benefit that reduces with income could support formal work incentives among low-income workers by compensating for modestly high average effective PIT rates and high and less progressive employee SSCs. The phase-in region of an EITC in-work benefit typically incentivises labor supply increases at the intensive margin (i.e. longer hours) as the tax credit amount increases with gross earnings.
The merits of an in-work benefit programme are stronger in individual-based tax systems such as in Lithuania. In an individual-based tax system such as in Lithuania, IWB work incentives are stronger relative to IWB in a household-based tax system (which depends on household size and is mean-tested) as it may reduce incentives for the second earner to enter work (Bargain and Orsini, 2004[15]).
In-work benefits are typically costly. Spending on the EITC programmes in the United States and the United Kingdom ranges from 0.4 to 0.5% of GDP (OECD, 2020[16]). However, these costs are at least partly outweighed by lower welfare payments since some jobseekers find a job as a result of the IWB. This in turn helps to broaden in the tax base. EITCs can be financed with PIT revenues rather than SSC revenues so that there is no loss to SSC funds.
A challenge for designing an EITC in Lithuania is that it could add to an already complex tax and benefit system. The complex structure of the EITC and its interaction with the rest of the tax and benefit system make its impact on labour supply dependent on its design and on household characteristics, such as marital status and number of children (Liebman, 1998[17]); (Eissa and Williamson Hoynes, 2004[18]). Estonia introduced a non-payable tax credit for low-income earners in 2016 and abolished it a year later in 2017, reflecting the challenges of adding additional complexity to the tax system. An EITC would increase marginal tax rates, which are already high for low income workers over the income range where the BA is tapered out.
Table 3.4. Selected advantages and limitations of in-work benefits
Source: OECD analysis. (Saez, 2002[19]) (OECD, 2005[11]) (Vandelannoote and Verbist, 2020[20]).
The in-work social assistance could be eased and the duration of assistance could be extended to improve work incentives
Lithuania operates in-work social assistance for individuals starting new jobs, but its current scope is limited. In-work social assistance (IWSA) is provided to out-of-work individuals that move into employment. The scope of IWSA is by definition narrower than in-work benefit, which is usually available to a broader group of low-income workers (only about 2 000 individuals currently receive IWSA). To be eligible for IWSA, taxpayers must meet several eligibility criteria including the following. First, recipients must be receiving SA benefits (i.e. qualify for the income means-test and other SA eligibility conditions for at least one month). Second, once in work, recipients must earn income of at least the MMW or the minimum hourly pay (thus part-time workers are included provided they earn at least the minimum hourly pay). IWSA is based on a person’s income before they enter work. It is not means-tested against current income. Third, recipients must be registered with the employment service for at least 6 months. The former two criteria are likely to significantly limit eligibility while the latter registration with the employment service is unlikely to disqualify many taxpayers (since it is a standard mandatory requirement for receiving SA by working-age adults).
In-work social assistance was reformed in 2020. To encourage social benefit recipients to enter the labour market, eligibility for IWSA was simplified in June 2020 by removing the condition on the upper wage limit, which was of approximately the AW. The amount of IWSA was doubled and the amount was set to decline with duration. For 1 to 3 months, the amount of IWSA is set at 100% of previous SA (paid during the last 6 months prior to employment). For 4 to 6 months and 7 to 12 months, the IWSA falls to 80% and 50% of the previous SA respectively. After 12 months, IWSA is no longer available.
The amount of IWSA depends on employment duration but not on current income. The reform in 2020 implies that the effect of IWSA on work incentives is the strongest at the beginning of employment spell. The PTR of entering work at the minimum wage is 27% with full IWSA (i.e. at one month of employment) and increases to 65% after 12 months of employment when the IWSA expires (Figure 3.15). Similarly, the PTR of entering work at the AW is 42% with full IWSA and 58% after IWSA expires (Figure 3.15). Since the IWSA upon entering work equals the previous SA out-of-work, initial IWSA removes all disincentives to work associated with the loss of means-tested SA. Work disincentives in the first 3 months of employment are produced only by SSCs and PIT.
Figure 3.15. In-work social assistance falls with longer employment duration
PTR for a single parent with two children starting work at the minimum and average wage, 2021
Source: OECD Tax and Benefit model TaxBen (version 2.4.1).
Redesigning the in-work social assistance with in-work benefit characteristics could improve work incentives but it would not be without challenges. Unlike an in-work benefit (IWB), the IWSA reduces with duration rather than with income (i.e. it is not means-tested with income), it has strict eligibility criteria and it is time-limited (Table 3.5). Some of the IWSA criteria could be redesigned to be more similar to an IWB to increase the incentives to work while also retaining its current SA function. However, such a redesign would not be without challenges. Setting the amount and taper rate of an IWB is perhaps the most challenging but important aspect of its design. Setting the amount depends on several factors including the difference between out-of-work income and in-work work income, the level of income support relative to the MMW, the cost of the programme (and how it could be funded) and the numbers of workers earning income at the income levels where the benefit is tapered. Additionally, sufficient time would need to be provided to induce wage progression given that most studies show wage progression at 3–4% per year (OECD, 2005[11]).
In-work social assistance could be increased beyond the level of social assistance similar to Slovenia to increase work incentives but this would come with several drawbacks. Several OECD countries have adopted social assistance-related measures that support incentives to enter work (Table 3.6). Slovenia’s Financial Social Assistance implies that those who work more hours have higher SA relative to those who work fewer hours (Table 3.6). This means a person entering work becomes eligible for higher SA, which is then tested against their current income. However, such a design has drawbacks as, apart from being relatively costly, it increases income at particular hour thresholds and may prevent work progression beyond these thresholds. A wider set of policies that increase work incentives for new employees in selected OECD countries is shown in Table 3.6.
In-work social assistance eligibility criteria could be eased and the duration of assistance could be extended to improve work incentives. Given the challenges of redesigning the IWSA so that it tapers with income like an IWB, a relatively more straightforward redesign option that has the potential to improve incentives to work include adjusting the eligibility criteria and time duration. IWSA could be broadened by reducing the requirement time to register with the employment services from 6 months to e.g. 3 months. The IWSA could be extended by increasing the current retention of 100% of the IWSA for 3 months for a longer duration such as 6 months.
Table 3.5. Comparing Lithuania’s in-work social assistance with an in-work benefit
Source: OECD analysis.
Table 3.6. Various policies that increase work incentive for new employees in selected OECD countries
Source: OECD TaxBen country reports available at http://oe.cd/TaxBEN.
The design of social security contributions
Most employer SSCs were shifted to the employee. With regard to SSC rates in 2019, employer SSC rates were cut (from 30.5% to 1.47%) and employee SSC rates were increased (from 9.0% to 19.5%) (Table 3.7). Consequently, total SSC rates (employer and employee) were cut by almost half in 2019. Between 2019 and 2021, SSC rates remain unchanged.
To compensate employees for higher employee SSCs, gross wages were adjusted upwards. The authorities increased gross wages mechanically by 28.9% in January 2019 in an attempt to keep the disposable incomes of employees unchanged. In 2018, the employer SSC rate was 30.5%, of which 28.9% was shifted to employees in 2019. The breakdown of the 28.9% included pension insurance of 22.3%, health insurance of 3%, and sickness insurance of 1.4% and maternity insurance of 2.2%.
Table 3.7. SSC rates in Lithuania, 2018 - 2021
Note: Employer SSC rates are comprised of 1.31% unemployment social insurance and 0.16% accidents at work and occupational diseases social insurance. Note that this is a general rate - In practice four categories of rates of insurance contributions for accidents at work and occupational diseases are applied, depending on company’s risk profile. Note that employer SSC ceiling has been abolished in 2021.
Source: OECD analysis.
Employer SSCs in Lithuania are comprised of three components.
First, the employer SSC rate in Lithuania of 1.47% in 2020 is comprised of:
1.31% for unemployment insurance, and
0.16% for insurance from accidents and work and occupational diseases.
Second, the employer payroll SSC rate of 0.32% in Lithuania is comprised of:
0.16% of gross wage for the Guarantee fund, and
0.16% of the gross wage for the Long-term employment benefit fund (to pay severance pay for the damage caused by occupational accidents or diseases).
Finally, an employer floor or minimum amount was introduced in 2018.
Lithuania does not levy SSCs on pensions, similar to about half of all OECD countries. PIT and SSCs play an important role in old-age income support. 21 out of 38 OECD countries levy SSCs on pensions (OECD, 2021[21]). Among OECD countries that levy SSCs on retirees, the SSC rate is always lower for retirees than for workers. In most OECD countries, pensions are not typically subject to pension SSCs or unemployment SSCs but they are sometimes subject to health SSCs (OECD, 2021[21]), which older individuals disproportionately use and benefit from.
The cuts to employer SSC rate and total SSC rates may have induced employers to raise employee wages to offset the decline in disposable incomes. Following the shifting of SSCs from employers to employees, employees are prima facie worse-off and employers better-off. The incidence of the change in the SSC mix generally falls on the employee (for a discussion of SSC incidence see Box 3.2) because there are few good alternatives to work for most people. Given that the evidence in the research literature for differential labour supply effects in employee and employer SSCs is scarce and weak (Box 3.2), it may be reasonable to assume that the labour supply effect of increases in employee SSC and employer SSCs are similar. Under this assumption, the labour supply effects of employee and employer SSCs can be usefully considered together (i.e. total SSCs). In Lithuania, total SSCs as a share of gross income have declined in Lithuania in recent years (Figure 3.16). Economic theory predicts that in the long-term (when the effective tax incidence becomes predominant) the incidence will largely fall on the employee through higher wages than would be the case otherwise. This occurs partly because employers must increase wages to offset the decline in disposable incomes for employees. The studies of (Gruber, 1995[6]) and (Anderson and Meyer, 1997[7]) show that employer SSC rate cut can lead to one-for-one increases in employee wages, which suggests that employer SSC incidence falls fully or mostly on workers. Average gross monthly earnings increased by 30% in Lithuania in Q1 2019 compared to the previous quarter, indicating that companies compensated workers with even higher wages than the 28.9% required by the upward mechanical adjustment. A similar wage increase in Romania following its shifting of SSCs from the employer to the employee (see Box 3.1).
SSC rates and revenues
Total SSC rates are somewhat low in international comparison but employee SSCs are among the highest in the OECD. Lithuania’s employee SSC rate of 19.5% is the third highest in the OECD, behind only Slovenia and Germany. The ratio of employee-to-employer SSCs in Lithuania is by far the highest in the OECD (Figure 3.16). Total SSC rates in 2020 are historically low in Lithuania. They are lower in 2020 than in any year going back to 2000. In addition, total SSC rates are lower than in Latvia and Estonia, where the total SSC rate is about 35% in both countries in 2020.
The shifting of virtually all employer SSCs to the employee in 2019 represented a significant departure from the SSC rate mix over the past two decades (Figure 3.16). As discussed previously in the PIT section (Figure 3.7), this shift was reflected in reduced SSC revenues (i.e. employee SSCs increased but were outweighed by employer SSC declines) and broadly revenue-neutral labour taxes (due to a compensating increase in PIT revenues). In 2020 however, while PIT revenues remain stable, total SSC revenues increased sharply to 8.7% of GDP from 8.1% in 2019. This increase moved Lithuania’s total SSC revenues from being below the OECD average to above it (total SSCs as a share of GDP are 8.4% in the OECD in 2019).
Figure 3.16. Total SSC rates are somewhat low in international comparison
Total SSC rates and tax wedge as a share of the average wage in 2020
Note: Analysis is for an individual at 100% of the AW.
Source: OECD tax database.
SSC ceilings
Since their introduction in 2019, SSC ceilings have been steadily cut. SSC ceilings were simultaneously introduced for employees and employers (excluding health insurance contributions) for incomes above 10 annual AW in 2019. The rationale for the SSC ceilings was partly to compensate for the higher PIT burden associated with the introduction of the progressive PIT rate system (discussed previously, see Table 3.2). In 2020, the employee and employer SSC ceilings (excluding insurance contributions) were reduced to 7 annual AW. In 2021, the ceiling for employee SSCs was further reduced to 5 annual AW and the employer SSC was abolished. From 2021, the employee SSC ceiling only applies to employees overall employment income (combined from all employers, as opposed to each employer individually as was previously the case).
Table 3.8. SSC ceilings have been steadily cut
SSC ceilings and top-PIT rate threshold, in annual AW, 2019 - 2021
Note: An SSC ceiling is the income level above which additional income is not subject to SSCs. Monthly AW are converted to annual AW by dividing by 12.
Source: OECD analysis.
SSC ceilings remain high in international comparison. The employee SSC ceiling of 5 times the annual AW in 2021 is high in international comparison. In OECD countries, the ceiling usually applies to wage levels higher than 1.67 of the AW (OECD, 2021[22]). Lithuania’s high employee SSC ceiling imply that high earners make a financial contribution to the funding of the social security system.
SSC ceiling cuts in recent years reduced the SSC burden on higher earners. The introduction of the SSC ceilings in 2019 reduced the SSC burden on high earners. Subsequent cuts to the SSC ceiling in 2020 and 2021 further reduced the SSC burden for higher earners. The abolition of the employer SSC ceiling in 2021 increased the SSC burden on higher income employers, albeit the increased SSC burden is small owing to the low employer SSC rate. The top-PIT rate threshold was also reduced jointly with the employee SSC ceiling between 2019 and 2021 (Table 3.8) and will have increased the PIT burden on middle and higher earners. The threshold values for both the PIT and SSCs is dependent on the size of the AW set by law each year which is in turn set based on actual wages in the economy.
The abolition of the employer SSC ceiling in 2021 increased the tax wedge on higher earners, but only modestly. The tax burden on high earners has increased with the abolition of the employer SSC ceiling in 2021 compared to when the ceiling was in place. However, the impact of abolishing the employer SSC ceiling on the tax wedge is limited by the low employer SSC rate and it only starts to take effect at high income levels (Figure 3.17).
Figure 3.17. The abolition of the employer SSC ceiling will modestly raise taxes on high earners
Tax wedge in Lithuania with and without the employer SSC ceiling, 50% - 800% of AW
Note: Compares employer SSC ceiling reduction to the same level as that of the employee SSC ceiling in 2021 (i.e. EUR 6 674 or 60 AW) with the abolition of the employer SSC in 2021. Given that the ceiling is adjusted annually to take account of wage changes, gross wage labour income is estimated by the OECD secretariat of EUR 16 426 in 2020 is used in the analysis (Taxing Wages 2021). A 2021 gross wage labour income could be estimated by increasing the figure by 9.9% (based on Statistics Lithuania wage growth) but this has little impact on the results. Source: OECD analysis; Taxing Wages 2021; Statistics Lithuania.
The merits of raising the employee SSC ceiling depends on several factors including the extent of SSC revenues that could be raised, arbitrage opportunities and equity considerations. For schemes such as health insurance, which are relevant given aging and relatively poor health outcomes in Lithuania (Figure 2.2), contribution ceilings imply that high-income earners do not participate with their full earnings in the redistributive scheme, which may not conform to the principle that people should contribute according to their financial capacity to do so. If Lithuania wished to generate additional resources to finance the social security system, particularly in healthcare where the benefit entitlements are mostly independent of the level of contributions paid, it could consider reinstating higher employee SSC contribution ceilings. The additional resources could help to reduce SSC contribution rates. The capacity for Lithuania to support income inequality by raising the employee SSC ceiling depends on several factors, some of which may need further empirical examination. The potential SSC revenue that could be raised by increasing the employee SSC ceiling will be limited by a relatively small number of high incomes (so revenue costing is needed). From an efficiency perspective, high-income earners could respond by leaving the country and working abroad. There may also be arbitrage opportunities for employees to switch to more lightly taxed income sources. From an equity perspective, raising the employee SSC ceiling weakens the link between SSC contributions paid and SSC benefits received. Raising the employee SSC ceiling may arguably be unequitable given that some higher earners will pay SSC contributions on a greater share of their income but receive no correspondingly higher pension SSC entitlement as a result.
Table 3.9. SSC and top-PIT rate threshold policy options
Source: OECD analysis.
High SSCs discourage work and encourage informality
High employee SSCs may reduce labour supply while low employer SSCs may increase labour demand. High employee SSCs may reduce labour supply and work incentives, particularly for individuals with weaker attachments to the labour market such as those on low incomes, older workers and second earners. High employee SSCs lower disposable income among low earners, potentially reducing work incentives. On the other hand, low employer SSCs may help to keep labour costs down for the employer and increase labour demand and to minimise distortions in firms and sectors where skills and labour productivity are low.
High SSCs could risk making informal employment relatively more attractive. In practice, employers are often in a position to transfer most of the SSC burden back to workers. Exceptions to this are when workers have strong bargaining power (which might be lower in an economy like Lithuania where there are large skill mismatches and relatively fewer high-skill workers in demand), when many workers are paid around the minimum wage (which cannot legally be reduced further), when workers value social protection (which may increase given visibility of employee SSCs) and when unemployment replacement rates are generous (not particularly generous in Lithuania). On this basis, employers may be able to transfer the SSC burden to Lithuania workers, albeit this is limited due to low employer SSCs.
The employer SSC floor ensures low-income workers receive a minimum SSC benefit contributions but it could risk disincentivising employers from hiring
The employer SSC floor was introduced in Lithuania in 2018. The employer SSC floor (or minimum amount), which was introduced in 2018, ensures that a minimum SSC contribution is paid on behalf of employees. For low-income employees with a monthly wage below the MMS (43% of AW in 2020), employers must calculate and pay employer and employee SSCs (at a total rate of 20.97%) on the MMS base for employees from an income base of at least the MMS. As such, employers pay employer and employee SSCs on behalf of employees based on the difference between the calculated wage and the MMS. The SSC floor does not apply to all workers and some are excluded including pensioners, disabled persons, young workers, workers receiving maternity, and paternity or child care benefits.
The employer SSC floor ensures that low-income workers make a minimum SSC contribution and indirectly ensures that they benefit from a minimum social benefit. At low income levels, SSCs would be low if it were not for the employer SSC floor due to relatively low employee SSCs, employer SSC and payroll rates (Figure 3.18). The employer SSC acts as a top-up by ensuring a minimum SSC contribution for low-income workers. The employer SSC floor also has the potential advantage of encourage employers to engage in the formal economy by helping to reduce the under-reporting of hours worked.
However, the employer SSC floor produces high tax wedges at very low incomes which could disincentivise employers from hiring low-income workers. At low-income levels (up to 21% of the AW in 2020), the employer SSC floor contribution is the largest contributor to total SSC contributions (at 10%, 15% and 20% of the AW, the SSC floor represents 72.6%, 64.3 and 52.9% of total SSCs) (Figure 3.18). As such, the employer SSC floor drives high SSCs and tax wedges as a share of gross incomes (at 10%, 15% and 20% of the AW, the employer SSC floor contributions represents 69.7%, 39.5% and 24.4% of gross incomes). This produces a perverse effect because the tax system makes it relatively more expensive for employers to hire low-income than high-income workers because the employer SSC is higher as a share of gross income at low incomes than at high incomes. If employers are responsive to the employer SSC component of labour cost at low incomes, this could reduce hiring at low incomes. At 44% of the AW, the employer SSC is floor phased-out and the effective employer SSC remains constant thereafter. Beyond 44% of the AW, total SSCs (employee and employer SSCs) continue to rise in nominal terms (because incomes are rising) but total SSCs as a share of gross incomes remains flat at 21.3% (19.5% + 1.47% employer SSCs + 0.3% payroll SSCs).
Figure 3.18. The employer SSC floor ensures that low income workers make a minimum SSC contribution
Decomposition of SSCs in Lithuania in 2020, 0 – 50% of the AW
Source: OECD analysis.
Whether the gap between the SSC floor and actual SSCs paid by vulnerable groups could be financed through general taxation depends on the extent to which the minimum wage is appropriately set and enforced. To encourage their employment, some vulnerable groups are exempt from the SSC floor (for example, parents that are receiving child-care benefits up until the child is 2 years of age, people with disabilities, those aged up to 24 years of age and the self-employed). The exemptions could contribute to these vulnerable groups receiving low social benefits (for example, illness benefits). Funding the gap between the SSC floor and actual SSCs paid up to the MMW for these groups through the state budget could help to support these groups. However, it could also encourage employers to try to pay full-time workers below the MMW if the MMW is set theoretically above worker productivity levels and MMW enforcement is relatively weak (i.e. if employers can pay workers below the MMW in the informal economy with limited risks of detection).
Employers appear to reduce hiring in response to the SSC floor
A simplistic comparison shows that there are fewer employees when the SSC floor applies compared to when it is not. Due to the employer SSC floor, employers pay employer and employee SSCs on an SSC base of the MMW for employees earning income below the MMW. The employer SSC floor produces a higher tax wedge than would otherwise be the case (in the absence of the SSC floor), which could disincentive employers from hiring low-income workers. However, do employers reduce hiring where the SSC floor is effective? Leaving aside the large share of employees earning the MMW, a simple comparison on either side of the MMW shows there are fewer employees for a given income range below the MMW than above it. 5.6% and 5.4% of employees earn between 10% and 20% and between 20% and 30% of the AW but a notably higher 8.0% and 7.2% earn between 50% and 60% and 60% and 70% (Figure 3.19, Panel A). This trend is consistent with reduced hiring below the SSC floor (although it does not demonstrate reduced hiring).
Employees that are exempt from the SSC floor are more concentrated where the SSC floor is in place. Employees that are exempt from the SSC floor include those receiving social insurance pensions or disability benefits and those under 24 years of age. These employee groups have greater distributional concentration below the MMW where the SSC floor is effective (Figure 3.19). At incomes below the MMW, the share of employees with a disability consistently exceeds the share of standard employees (44% vs 26%) whereas above the MMW (i.e. between 50% and 60% of AW) the former is only modestly above the latter (21% vs 18%) (Figure 3.19, Panel B). A similar trend is observed for employees receiving an SSC pension who are also exempt from the SSC floor (Figure 3.19, Panel B). Below the MMW, the share of employees aged under 24 years far exceeds the share of standard employees (67% vs 26%). However, it is possible that these employees are more concentrated in this part of the income distribution for reasons other than the operation of the SSC floor such as they tend to earn lower incomes for different reasons due to having lower skills and educational attainment (which could be controlled for in an econometric setting as part of future work).
Employers hire employees just below the SSC floor exemption age cut-off at a greater rate than employees just above the age cut-off. The SSC floor age cut-off (i.e. exempt under 24 years of age) facilitates a quasi-experimental comparison of the share of employees by income just above and below the age cut-off. Employers should prefer to hire a 23-year-old who would be exempt from the SSC floor and attracts a lower tax wedge than a 24-year-old to whom the SSC floor applies (holding other factors constant). Below the MMW where the SSC floor applies, the share of employees aged 22 and 23 (i.e. 46%) is 56% higher than the share of employees aged 24 and 25 (i.e. 29%) (Figure 3.19, Panel D). This compares to just 4% higher above the cut-off (i.e. 40 – 60% of AW). This provides indicative evidence that employers hire employees that are exempt from the SSC floor at a greater rate than those that are not. A more precise estimate could be obtained by controlling for additional explanatory variables such as skills and educational attainment in a regression discontinuity design as part of future work.
Figure 3.19. Some evidence is consistent with employers reducing hiring where the SSC floor applies
Employee groups that are exempt from the SSC minimum floor, % of employees, by employment income as % of AW
Note: Analysis based on 43,389 individual taxpayers with positive employment income. Of these, 5.2% receive disability benefits. 8.3% are aged less than 24 years of age. 8.1% receive an SSC pension. 3.2% and 3.5% are aged between 22 and 23 and 24 and 25 respectively. Two ages are to increase sample size.
Source: OECD analysis of microdata 2019.
To support better health outcomes and inclusive growth, Lithuania could raise social spending funded by SSCs and general taxation
Low levels of social spending could be increased to support inclusivity. Social spending as a share of GDP in Lithuania is low in international comparison (Figure 2.19). Increasing social spending would help to ensure that the incomes of those who are not benefiting from rapidly rising wages (Figure 2.7) continue to see their standards of living increase (e.g. the unemployed, inactive workers and individuals with health issues).
Health expenditure is primarily financed through compulsory health insurance. In Lithuania in 2019, health expenditures are primarily financed through compulsory health insurance (58% of all financing) and out-of-pocket spending (32%) and much less so by government schemes (8%) and voluntary health insurance (1%).6 By comparison, in the OECD on average in 2019, compulsory health insurance represents 39% of expenditure followed by 35% in government schemes and 20% out-of-pocket expenditure (OECD, 2021[23]). Therefore, compulsory health insurance represents a relatively more important share of health expenditure in Lithuania than the OECD more generally.
Compulsory health insurance in Lithuania relies mainly on SSC funding and to a lesser extent on transfers from government. Compulsory health insurance, the primary source of finance for health expenditure in Lithuania, is primarily financed by SSCs (i.e. 69% of total financing in 2019) (Figure 3.20). However, the pension and health funds are likely to face constraints which are expected to worsen given a rapidly ageing population and the associated increased costs associated with pensions (Figure 2.2).
Figure 3.20. Lithuania’s welfare system relies mainly on funding through SSCs
Financing sources of compulsory health insurance by type of revenue, OECD countries, 2019 (or nearest year)
Note: Other includes compulsory prepayments and other domestic revenues.
Source: (OECD, 2021[23]).
SSCs in Lithuania are currently focused towards supporting pensions and health. Of the employer and employee SSC rates, the largest shares are allocated to pension, health, sickness, maternity and unemployment. The employee SSC rate of 19.5% is divided into: pension (8.72%), health (6.98%) sickness (2.90%), and maternity (1.71%). The employer SSC rate of 1.79% is divided into: unemployment (1.31%), accidents at work and occupational diseases (0.16%) and other contributions to long-term-work benefit and the guarantee fund (0.32%).
In 2019, Lithuania partly shifted the funding of social insurance pensions from SSCs to general taxation, but the reform may have limited impact on the general financial sustainability of the social insurance fund. Lithuania’s shifting of the financing of the non-earnings-related part of social insurance pensions to general taxation will not materially change the overall resources for social expenditures. In 2019, the authorities shifted the financing of the general part of pension SSCs to general taxation and to the PIT. The employee pension SSC rate was 8.72% in 2019 (where it currently remains in 2021), which represented a reduction from the previous employee pension SSC rate of 3.0% plus the employer pension SSC rate of 22.3% in 2018. To offset the reduction in the pension SSC rate, the progressive PIT rate was introduced and part of the PIT revenue is used to fund social insurance pensions. However, this shift to funding social insurance through the PIT represents only a substitute for the lost SSC funds and does not materially increase resources to the SSC fund.
The sources of financing to the welfare system could be diversified. The welfare system relies heavily on SSC for financing and to a lesser extent on transfers from government in international comparison. Healthcare funding had previously been shifted from taxes to SSCs. However, financing through the SSC system may be challenging, as it requires increasing already high SSC rates on low incomes. A more balanced financing mix may be optimal to prevent a drop in pension benefits given rapid aging and poor health outcomes. Financing the welfare system through general taxation has some advantages as well as limitations (Table 3.10). Financing social benefits through general taxation instead of SSCs can reduce the tax burden on labour income, particularly if social benefits are financed through taxes that do not bear solely on labour income. Reducing the tax burden on labour income through lower employer and employee SSCs can in turn provide greater incentives for employers to hire workers and for workers to participate in the labour market.
Alternative financing options for maternity leave and childcare benefits exist. Women can take maternity leave starting the 30th week of pregnancy. Since the rationale for maternity leave is that women are unable to work due to their health status (i.e. similar to sick leave), it could be funded by SSCs as it is an insurable event. Parents can take childcare benefit up to 2 years after a child is born. Childcare is paid to compensate for childcare costs and lost employment. Lithuania has argued that the former could be paid for by a basic flat amount from general taxation and the latter from SSCs. Following the same principle, a further possibility could be to finance childcare benefits and allowable leave through general taxation. If additional leave is used beyond the legal allowable limit, it could be financed with SSCs. Maternity benefits are often funded mainly through general taxation in EU countries.
In a changing world of work, financing social benefits partly through general taxation could help ensure that welfare support remains available for a large number of people. Structural changes in the economy including digitalisation and automation are resulting in an increasing number of workers which make smaller SSC contributes (self-employed, temporary workers and workers with irregular working hours (OECD, 2017[24]). These structural changes in the economy present new sustainability challenges for welfare systems that are financed primarily through SSC contributions.
Table 3.10. Advantages and limitations of financing health systems
Source: International labour organisation. Adapted from (OECD, 2018[25]).
The design of the benefit system
The design of unemployment benefits
The rationale for unemployment benefit programmes in most countries include protecting workers against joblessness while containing poverty. Unemployment benefit (UB) programmes protect workers against the risk of income loss during joblessness. They also reduce the risk of wage losses upon re-employment by allowing workers to spend more time to find a job that matches their skills than they might otherwise. Unemployment insurance (UI) programmes help to contain poverty and to ensure both a fairer distribution of income in society (OECD, 2020[26]).
Unemployment benefit coverage is limited by strict entitlement conditions. Factors that determine UB in Lithuania include the contribution record, the minimum and maximum amounts, the fixed and variable components and the benefit duration. With regard to eligibility, unemployment benefit is available to people aged over 16 who have not reached retirement in Lithuania. SSCs must have been made for at least 12 months in the last 30 months, which has now been brought in line with the 12 month median minimum contributions period across OECD countries (OECD, 2018[27]). However, only about a third of the unemployed registered with the Employment Services are entitled to unemployment benefit (Užgalė et al., 2020[28]). While strict entitlement conditions increase the financial incentives to find employment quickly, they also bear the risk of strongly penalising vulnerable population groups who struggle to find employment despite job search efforts. The unemployment benefit is neither means-tested nor taxable and unemployed persons can re-apply for unemployment benefit after 12 months if the minimum contribution requirements are met (note that unemployment benefit is included in the means-test for social assistance benefit).
The unemployment benefit consist of fixed and variable components, set by the authorities each year. The fixed component set at 23.27% of the MMS (in the month for which the UB is paid) produces an UB floor. The minimum UB is EUR 149.39 in 2021, up modestly from EUR 141.25 in 2020 and based on the fixed component (at a rate of 23.27%).7 The max UB of EUR 840 in January 2021 is up from EUR 760.01 in 2020. The max is set at 58.18% of the gross monthly AW as calculated by the Lithuanian Department of Statistics.8 The variable component, which is linked to the former level of insured income of the unemployed,9 is 38.79%, 31.03% and 23.27% in months 1 to 3, 3 to 6 and 7 to 9 respectively.
Similar to some OECD countries, the unemployment benefit design in Lithuania is determined by previous income, time-limited and declines with unemployment duration. In most OECD countries initial income support for the unemployed is provided through contributory unemployment insurance systems in which benefits for eligible job losers are a function of the previous wage and are time-limited (OECD, 2020[26]). Among countries with declining UB schedules like Lithuania, Chile, Latvia and Slovenia, the initial net replacement rate (NRR) are often high (OECD, 2020[26]). Unemployment NRRs show the proportion of the net household income before the job loss that is kept (note that since the OECD TaxBen NRR indicators assume a full contribution record) (Figure 6.2). In all countries, the transition from UB to social assistance benefit produces a decline in the NRRs. Hence, even in countries with constant UBs such as Austria, France, Germany and Portugal, the level of support for the unemployed declines over time. Lower UBs in the long-term imply lower income support at a time when workers might value it the most as they progressively deplete any existing assets to smooth consumption (OECD, 2020[26]).
Lithuania’s unemployment duration of 9 months is of a relatively short duration in international comparison. The duration of UB is 9 months in Lithuania regardless of contribution history. For unemployed persons not more than 5 years below the retirement age, the duration of the UB is extended to 11 months (provided early retirement is not available to them). The rationale for reduced UB over time in Lithuania is to provide higher levels of income stabilisation during the initial months of unemployment and higher incentives to take up a job thereafter (Navicke, Avram and Demmou, 2016[29]). In most OECD countries, UBs are granted for at least twelve months (Koutsogeorgopoulou, 2020[30]). The maximum duration of UB varies from less than six months in Hungary and Israel to close to three years in Iceland and Sweden (OECD, 2020[26]).
At low and middle incomes, the unemployment benefit design produces greater incentives to find work for longer unemployment spells. For longer unemployment spells, unemployment benefit is cut producing lower NRRs (Figure 3.21). For low, middle and high incomes (at 50%, 100% and 150% of the AW respectively) at short unemployment spells (1 – 3 months), the NRR declines from 83% to 76% to 59%. For medium unemployment spells (3 – 6 months), the NRR starts lower at 72% and falls to 64% and 59% and for longer unemployment spells (6 – 9 months), the NRR starts lower again and falls to 52% and 49%. The declining variable rate means that for longer unemployment spells, higher levels of previous (insured) income are needed to reach the same max UB. For short, medium and longer UB spells, the maximum UB is reached at gross incomes of 114%, 142% and 190% of the AW.
At high incomes, unemployment replacement rates are lower overall but remain constant for the unemployment period before declining sharply thereafter. At and above 190% of the AW, the UB is constant even with longer unemployment spells (Figure 3.21). A worker with previous insured income of 200% of the AW receives the same UB of EUR 760 (45% of the AW) throughout the 9 month duration. For these high earners, the subsequent decline in income as they transition from UB to social assistance benefits is relatively greater compared to lower earners thus producing one very large incentive to find work before the 9 month mark.
In countries such as Lithuania where unemployment benefits are untaxed, net replacement rates tend to be higher. UBs are not taxed in Lithuania, unlike several OECD countries such as Belgium, Canada, Denmark, Ireland and Norway. Any increase in labour taxes reduces in-work income without reducing out-of-work income, thus increasing NRRs and the attractiveness of unemployment in addition to the relative bargaining power of employees over employers.
In international context, low-income unemployed in Lithuania have relatively high NRRs for short and medium unemployment spells but much lower NRRs for long unemployment spells. For short spells of unemployment, low-income singles in Lithuania have among the highest NRRs in the OECD (82.5% vs 67.1%) (Figure 3.21). For medium spells, low-income singles in Lithuania have NRRs similar to but modestly above the OECD average (60.8% vs 52.9%). Therefore, NRRs are relatively high in Lithuania during the current 9 month UB duration. For longer spells of 10 months, NRRs in Lithuania are among the lowest in the OECD (because the UB duration of 9 months has expired and unemployed persons have transitioned to social assistance benefits). Consequently, under the three unemployment periods examined for a low-income worker, the NRRs in Lithuania change from among the highest NRRs after 2 months to among the lowest NRRs after 10 months. Given Lithuania’s UB duration of 9 months, the sharp decline in the NRR at 10 months is significant relative to other countries. Given UBs in Lithuania are mostly determined by previous incomes, and that wages are somewhat low in Lithuania and include an income cliff at older ages (see Figure 2.7 and Figure 2.12), unemployment NRRs in Lithuania should remain relatively high to prevent the unemployed from falling deeper into poverty.
Figure 3.21. The unemployment benefit design produces greater incentives to work for longer unemployment spells
Unemployment benefit (fixed and variable portions) and NRR, by previous gross wage as a share of AW, 0 – 200%
Note: The net replacement rate (NRR) shows total unemployment benefit as a share of net income (which represents a proxy for previously insured income in the model). Unemployment NRRs are calculated by comparing net family income out-of-work with net family income in work. Due to very high NRRs at low income levels, the NRR starts at 15% of AW. High NRR below this threshold represent a very small proportion of individuals.
Source: OECD analysis.
Figure 3.22. In international comparison, unemployment replacement rates in Lithuania are high for short spells but low for long spells
Net replacement rate in Lithuania and OECD countries, single individual at 50% of the AW, after 2, 9 and 10 months of unemployment, in 2020
Note: OECD average is an unweighted average of the countries shown. OECD TaxBen NRR indicators assume a full contribution record.
Source: OECD Tax and Benefit model TaxBen.
Poverty risks are high for minimum wage Lithuanian workers who become unemployed. Lithuanian workers with previous gross earnings of 60% of AW or less in 2021 would be below the poverty line in the first three months of unemployment (Figure 3.23). A worker earning the minimum monthly wage in 2021 (46% of AW) would thus be further below the poverty line in unemployment. A distributional question then arises as to the share of Lithuanian workers that earn below these income thresholds who would face poverty risk in the event of unemployment. Given that 40% of individuals have monthly gross wage incomes of between EUR 500 and EUR 1,000 in 2020 (i.e. 35 – 70% of AW as calculated using the Lithuanian AW), the cohort who would face poverty in the event of unemployment is quite large. More generally, 56% of the unemployed are at risk of poverty in Lithuania 2020, up 2 percentage points on 2019. On the other hand, low-income unemployed persons in the first three months of unemployment have relatively high unemployment net replacement rates (NRR). For single individuals earning half the AW, NRRs are 83% in Lithuania vs 67% in the OECD, albeit they decline sharply in later months.
Figure 3.23. Poverty risks are high for minimum wage Lithuanian workers who become unemployed
Unemployment benefit in the first 3 months of unemployment, by previous gross income, 2021
Note: Long and continuous employment record (22 years) is assumed. Based on the 2nd month of unemployment. At a previous income at the minimum monthly wage, unemployment benefit is EUR 398 (fixed = EUR 642 x 23.27% and variable = EUR 642 x 38.79%).
Source: OECD Tax and Benefit model TaxBen.
Reform options
Given comparatively high levels of long-term unemployment and to give workers more time to find a job that matches their skills, the duration of unemployment benefit could be extended. Long-term unemployment in Lithuania (10 months or more) is relatively high in Lithuania compared to the OECD and over 20% of jobseekers stayed on unemployment for at least 9 months in 2016 (OECD, 2018[27]). To support the long-term unemployed, the maximum period of UB could be extended, as has been proposed in previous OECD Economic Surveys (OECD, 2020[31]) (OECD, 2018[32]) and brought more closely in-line with the OECD average (of 12 months). Given high skill-mismatches in Lithuania, an UB extension would provide the unemployed with more time to find a job that matches their skills. This underlines the importance of implementing active labour market policies. From a distributional perspective, extending the UB duration would benefit those at relatively higher poverty risk such as the low-skilled and older-workers particularly elderly women since they are more likely to be found among the long-term unemployed.
Alternatively, a targeted unemployment benefit duration beyond one year could be aimed at pre-retirement age individuals. The unemployment benefit duration for pre-retirement age individuals (55 – 64 years) is currently 11 months compared to 9 months in the general population. However, this cohort are more likely to be unemployed for longer durations, reflecting additional challenges faced by older workers to re-enter employment including weaker ICT skills, skill mismatches and poor health. Extending unemployment benefit duration will give them more time to find a job rather than transferring to disability and early-retirement pension schemes. Extensions to the unemployment benefit duration should be designed so that the unemployed do not risk falling into unemployment at later unemployment spells.
The extension of the unemployment benefit duration should not be financed through reduced net replacement rates. The extension should not be financed through reducing the relatively high NRRs in the initial months of unemployment and reducing the UB income cap, which affect those on (previously insured) high incomes more. The former option would risk pushing the unemployed further into poverty. While the latter option would make the UB design more redistributive and equal, it would weaken the link between incomes and UBs which encourages formal work. These options could be considered in the long-term if wages were to increase significantly in Lithuania. Higher wages if achievable in the long-term would also support the currently low levels of social expenditure through higher labour taxes so that the duration of UB could be extended while maintaining the design.
To contain the high at-risk-of-poverty rates among the unemployed, eligibility criteria could be eased to widen unemployment benefit coverage. Under the New Social Model in Lithuania in 2017, the coverage of UBs were increased by easing eligibility conditions and made more generous by extending the duration of benefits to 9 months and raising payment rates (OECD, 2020[31]). Eligibility for unemployment benefit was eased with the minimum SSC contribution period cut to 12 months, in line with the OECD median period (some OECD countries provide up to 3 years). However, given high and rising poverty rates among the unemployed (more than half the unemployed are at risk of poverty in 2020) (Figure 2.11), and that a low one-third of the unemployed registered with the Employment Services are entitled to UB (Užgalė et al., 2020[28]), coverage could be further extended by easing eligibility restrictions. However, the impact of reducing the minimum SSC contribution on poverty and UB coverage may be limited (MoSSL impact assessment).
Lithuania is considering an automatic adjustment of minimum SSC contributions for unemployment benefit from 12 to 6 months during periods when the government declares an economic crisis. However, this could raise definitional and design challenges. Economic crisis would have to be carefully defined to determine when the automatic adjustment is triggered. Different types of economic crisis would also affect different workers differently and their relative need for reduced SSC contributions.
Poverty risk faced by the unemployed is linked to relatively low wages. As we have seen UBs can be high relative to wages for low and middle-incomes, but they are also low relative to wages in other OECD countries because they are by definition mechanically linked to Lithuanian wages since the fixed portion of UBs are anchored against the government set level of MMS (which is set annually according to wage levels) and the variable portion of UBs is based on previous income levels. Even if Lithuania were to increase the fixed and variable rates used to calculate UBs, wage would remain a key determinant of UB.
Lithuania should maintain a strong link between previous income and unemployment benefits to encourage formal work. The variable component of UB is linked to the former level of insured income, which raises the relative attractiveness of formal employment. A strong link between previous earnings and UB encourages workers to progress in formal work by increasing income, effort and hours and more broadly investment in a career in the formal economy because, in the event of unemployment, they will be financially additionally protected through higher levels of UB. However, as this link weakens for higher earners as the maximum limit on UB is imposed, formal employment then becomes relatively less attractive for higher earners.
Lithuania should continue its policy of not taxing unemployment benefits. As discussed previously, the Lithuanian authorities have decided against taxing UBs, unlike several OECD countries (such as Belgium, Canada, Denmark, Ireland and Norway). However, since SSCs are non-deductible from taxable personal income in Lithuania (i.e. they are paid out of after tax earnings), also unlike most OECD countries, there is little justification for taxing UBs in Lithuania. In addition, taxing UBs would further reduce NRRs risking pushing individuals further into poverty.
The design of housing benefits
Housing assets are a main source of wealth in Lithuania. In a paper on housing wealth using microdata in Lithuania, (Bielskis and Ciginas, 2020[33]) make a number of important observations. First, the home ownership rate at 93.2% in Lithuania is among the highest in the EU. Second, total household assets in Lithuania are comprised of mostly of real assets (95.3%). Of all real assets, 67.5% is accumulated through housing as the main residence with a further 18.2% related to other real-estate10. Accordingly, the share of renters is low and the lowest real estate wealth is accumulated by renters. Third, the authors distinguish between households that own old Soviet-era houses without mortgages but their wealth value from real-estate is small compared with younger households that own new houses with mortgages that have higher net wealth value. Indeed, a high 51.3% share of low-income households in 2017 in Lithuania live in dwellings that are owned outright (OECD, 2018[32]). This means the value of real estate can be relatively high even for some of the lowest income households (i.e. individuals that hold little liquid wealth but sizable illiquid assets) (OECD, 2017[24]).
However, housing quality is relatively poor. Around 6.5% of the low-income population (bottom quantile) in Lithuania experience severe housing deprivation, living in overcrowded and poorly equipped dwellings (OECD, 2018[32]). This figure is high in international comparison. Housing quality remains an important challenge in post-socialist countries where much of the stock of housing was privatised after the independence; the current owners of these dwellings often have financial difficulties to maintain and upgrade them (OECD, 2017[24]).
Lithuania has a generous housing benefit but its reach is limited. Lithuania has among the most generous benefit schemes in the OECD (Figure 3.24, Panel A). For families earning 10% and 50% of the AW, rent allowance represents 33% and 17% of gross wage respectively. However, the reach of the housing benefit is limited (Figure 3.24, Panel B). Of families in the bottom quintile, only 9% receive the housing allowance.
Figure 3.24. Lithuania has among the most generous housing benefit schemes in the OECD but its reach is limited
Note: Panel A: 1. Rent allowance calculated based on assumed rent of 20% of average wage. 2. Only shows central government housing allowance. Where no national scheme exists, a representative region was chosen, refer to country specific information for more details: http://www.oecd.org/els/soc/benefits-and-wages-country-specific-information.htm. 3. Full-time earnings are either at the 10th or the 50th percentile of the full-time wage distribution. No transitional benefits for entering the labour market are considered; social assistance but no unemployment benefits are considered. 4. The four family types considered are (1) single person, (2) single parent with two children aged 4 and 6, (3) one-earner couple and (4) one-earner couple with two children aged 4 and 6. Earnings are either at the 10th- or the 50th-percentile of the full-time wage distribution. 5. Data for New Zealand are preliminary and data for Korea refer to 2018. Panel B: 1. No information available for Australia, Canada, Chile, Colombia, Costa Rica, Japan, Korea, Mexico, New Zealand, Türkiye and the United States due to data limitations. Only estimates for 100 or more data points are shown. 2. Quintiles are based on the equivalised disposable income distribution. Low-income households are households in the bottom quintile of the net income distribution. 3. Data for Germany and Italy refer to 2019 and for Iceland to 2018. 4. In the United Kingdom, net income is not adjusted for local council taxes and housing benefits due to data limitations.
Source: (OECD, 2021[34]), indicator PH3.3.
Housing benefits aimed at renters and those in poor quality homes is significant, particularly for larger high-income families
One type of housing benefit in Lithuania that partly covers rental costs, available to renters and those living in worn-out dwellings, can represent a significant share of income. Housing benefit policy in Lithuania represents a large and complex policy area. This section focuses on the design and incentives produced by one type of housing benefit (HB hereafter) (būsto nuomos mokesčio dalies kompensacija) modelled in OECD TaxBen. Under this HB, part of housing rental costs are reimbursed by individuals and families under certain conditions including if they do not own a property, if the property is sufficiently worn-out or small in size. The HB is means-tested based on net income by family type whereby larger families (either with more adults or children) can continue to receive the HB at higher income levels11. While the HB remains unaffected below the limit, it ceases in-full and at once when the limit is exceeded. For eligibility, assets and income cannot exceed defined limits. The HB is also capped at the level of rent (depending on the region and other factors), assumed at 20% of gross income in OECD TaxBen. HB per person is equal to the amount of compensation multiplied by the number of family members. In 2020, the HB is set at EUR 92.92 per person (7% of AW). A couple or a single individual with one child (i.e. both have same family size) have a max HB of EUR 185 (14% of AW) while a couple with one child have a max HB of EUR 274 (20% of AW).
Housing benefits are larger for larger families but are capped at rent levels. The value of the HB is substantial. For a single person, a parent with two children and a parent with 3 children, the HB represents 7%, 20% and 20% of the AW respectively (Figure 3.25). For a parent with 2 children, the HB value is 2.3 times that of child benefits. For larger families, HBs are greater and remain in place at higher incomes (Figure 3.26).
The financing of the housing benefit is partly shifting to the municipalities. A reform in December 2021 transfers a portion of the financial responsibility of the housing benefit scheme from the State to the municipalities in January 2024.
Figure 3.25. Housing benefits are larger for larger families but are capped at rent levels
Housing benefits as a share of the average wage in Lithuania, by family type, 2020
Note: The HB is also capped at the level of rent which in OECD TaxBen is assumed to be 20% of gross income.
Source: OECD Tax and Benefit model TaxBen.
Figure 3.26. Housing benefits remain in place at higher incomes for larger families
Housing benefits as a share of the average wage in Lithuania in 2020, for different family types, by gross income
Source: OECD Tax and Benefit model TaxBen.
Housing policy options
Consideration could be given to expanding the availability of the housing benefit. Lithuania has one of the most generous HB schemes in the OECD but its reach is limited (Figure 3.24). Expanding the availability of the housing benefit could be considered.
The availability of the housing benefit could be expanded, perhaps in particular to support single individuals. Single individuals are among the most at risk of poverty groups in Lithuania (Figure 2.11). From the perspective of household size, single-person households are the poorest (Bielskis and Ciginas, 2020[33]). Compared with larger families with two adults or with children, single individuals receive not only relatively smaller housing benefits but those benefits are fully withdrawn at a lower level of income than for larger family types (approximately 80% of the AW), implying that middle-income single individuals just below the AW could also be at risk. Single individuals are less likely to own property (Bielskis and Ciginas, 2020[33]) and may be more likely to be renters. Increasing the income level at which HB is available could be considered to support middle-income families but it would be costly and the withdrawal threshold is relatively high in international comparison. A preferred option could be to expand the availability of the HB with a focus on single individuals.
The housing benefit could be reviewed and modified if those who benefit face limited poverty risks. The housing benefit is not widely used in Lithuania but the number of individuals benefiting from it has grown steadily. The housing benefit is an alternative to renting social housing. The housing benefit does not appear to be particularly well targeted.
Consider evaluating the effects on income inequality of the real estate tax including the relatively high exemption limit. Lithuania has among the highest rates of homeownership in the EU. The value of housing assets is relatively high and concentrated at the top of the income distribution. The progressive real estate tax is designed to increase the tax burden on households with larger housing wealth assets. Households with smaller housing wealth assets are exempt from the tax (up to EUR 150 000). The exemption is relatively high compared to other OECD countries. The reform has the potential to raise relatively low property tax revenues in Lithuania. Recurrent taxes on immovable property, in particular when owned by households, are the least damaging tax to long-run economic growth. A possible challenge with this reform that could be considered further is whether the tax could excessively burden groups that, despite owning property, have low incomes and face poverty risk. The rental market in Lithuania appears to be thin and informal. There are no tax deductions on the income generated through rental activities.
If income bunching is detected where the housing benefit is withdrawn, the feasibility of tapering more gradually could be considered. The HB is withdrawn fully and immediately producing high METRs for single individuals who may then face poverty traps and who already face high poverty risks. There are also high METRs for two adult households with three children who face high poverty risks and two adult couples with two children who face moderately high poverty risks. If income bunching were detected, the administrative feasibility of tapering the housing benefit withdrawal could be considered.
The design of social assistance benefits
Social assistance benefit in Lithuania provides last resort income support for low-income households in need. Social assistance benefit (SA hereafter) is a last resort income support available to low-income individuals and families in need in Lithuania. A small 2 to 3% of the Lithuanian population receive SA. SA is available to a range of low-income individuals (and families) including the unemployed, retirees, low-income workers, students, persons taking care of children and other dependents. SA is granted for 3 months, but it can be renewed an unlimited number of times if circumstances remain unchanged. For unemployed persons of working age that are capable of work, the amount is progressively reduced with duration.12 Since 2021, the benefit for a single person decreases in general following a different reduction schedule.
Social assistance is based on state-supported income, which is linked to consumption needs and several other interconnected policy levers. SA is determined by several policy levers set by government. The factors are formalised mathematically in equation (1). The base of SA is represented by state-supported income (SSI). In terms of the calculation of SSI and shown in equation (1), the amount of SA in 2020 is calculated as the difference between SSI and income (defined as the average monthly net income of the family together per person13) so that the SA increases for each additional person in the family but at a declining rate (this occurs through the functioning of the equivalence scale). For example, for a single person in Lithuania in 2020, SA is 100% of the difference between the SSI and the average net monthly income per single person (where net income refers to gross income less PIT and SSCs). For persons living together and the second and third person, the share is reduced for each additional person, which is referred to as the income equivalence scale.14 Other factors in equation (1) that increase the level of income that can be disregarded for means-testing the SA (i.e. income disregards) and family size. The income disregards (ID) rule makes work more attractive by ‘disregarding’ a portion of net income from the SA means-test, conditional on the family structure. Overall, the authorities can then increase the level of SA in several ways including by raising the level of SSI, the income equivalence scales per person and income disregards. SA will also be higher for larger families (in equation 1 both directly through the numbers in the family and also more indirectly through a higher cumulative income equivalence scale) and for lower net income levels.
(1) SA = [SSI – (net income/no. family members)] x [cumulative income equivalence scale]
The indexation mechanism of the SSI could be improved. The SSI is indexed in autumn in year T (to prepare for the state budget) based on the amount of minimum consumption needs (AMCN) calculated in year T–2 using inflation forecasts for year T-1 ) (Navicke, Avram and Demmou, 2016[29]). This time lag has resulted in recent increases in social benefit not keeping pace with rising inflation.
State-supported income was further increased in 2022. In January 2020, 2021 and 2022, SSI was increased from EUR 125 to EUR 128 to EUR 129 respectively. In June 2022, it was further increased to EUR 147 to combat rising energy prices.
The amount of social assistance was increased and differentiated by payment duration in June 2020. For a single person, when SA is paid up to 6 months, the amount of SA is the difference between 1.4 SSI per person per month and the actual income of a single person. When SA is paid up to 6 to 12 months, SA is the difference between 1.2 SSI per person per month and the actual income of a single person. When SA is paid for more than a year, SA is the difference between 1.1 SSI per person per month and the actual income of a single person. With regard to a family, for a first family member, SA is the difference between the amount of 1.1 SSI per person per month and the actual income of the family member. For a second family member, SA is 90% of the difference between the amount of 1.1 SSI per person per month and the actual income of the family member. For a third and subsequent family members, SA is 70% of the difference between the amount of 1.1 SSI per person per month and the actual income of the family member.
However, social assistance benefits in Lithuania reduce poverty by less than in the OECD average. Figure 3.27, Panel A shows income of a jobless single individual claiming guaranteed minimum income benefits (but not unemployment benefits) as a percentage of median disposable income in Lithuania and OECD countries. The analysis highlights that minimum incomes in Lithuania alleviate poverty risks to a lesser extent than the OECD average for jobless single individuals, suggesting that Lithuania could consider increasing the SSI for this group.
Minimum income supports have increased recently. Minimum-income benefits form part of the redistribution system. Poverty avoidance and poverty alleviation are primary objectives of minimum-income benefits. An indicator of income adequacy (IA) is the guaranteed minimum income (GMI) as a share of the median disposable income. GMI is the net income of a jobless family claiming social assistance benefits that is not eligible for unemployment benefits (as they have expired). Lithuania has increased minimum incomes in recent years, more so for jobless persons with 2 children than for persons without children (Figure 3.27, Panel B).
Figure 3.27. Income adequacy improved in 2021
Guaranteed minimum income amount as % of median disposable income, Lithuania and the OECD, 2005 – 2021
Note: Housing benefits not included. Both the numerator and denominator of the income adequacy indicator are adjusted for family size using the square root of the family size.
Source: OECD Tax and Benefit model TaxBen.
Benefits of last resort are modest for single individuals but among the highest for families with children. GMI as a share of median incomes are below the poverty line in Lithuania and in most OECD countries (Figure 3.28). A single individual without children in Lithuania has a minimum income of 29% of the median disposable income in 2021 vs a poverty line of 50% of the disposable median income. GMI as a share of median incomes in families with children is higher relative to single persons in Lithuania and in OECD countries generally.
Figure 3.28. Income adequacy is relatively high for families with children
Guaranteed minimum income amount as % of median disposable income, 2021
Note: Housing benefits not included. Both the numerator and denominator of the income adequacy indicator are adjusted for family size using the square root of the family size.
Source: OECD Tax and Benefit model TaxBen.
Social assistance increases progressively for larger size families. As implied by equation (1), the max SA benefit increases progressively with larger family size (Figure 3.29).15 In 2020 the max SA for a single person would be EUR 125, a couple EUR 225 (EUR 125 + EUR 125*80%), a single parent with two children EUR 313 (EUR 125 + EUR 125*80% + EUR 125*70%) and a single parent with three children EUR 400 (and so on). In terms of the rate of increase of the SA, since the income equivalence scale in 2020 declines with additional persons (100% of the difference for the first adult, 80% for second and 70% thereafter), after the second person, the 70% share remains constant and the rate of SA increase remains the same (in absolute terms). In addition, social assistance declines as incomes increase from low to middle-income levels, at first quickly and then more gradually as incomes rise for different family types.
On top of social assistance, a further pupil assistance is available to low-income families with school children. On top of SA, pupil assistance is also available to low-income families with school-age children. The amount of the pupil assistance is BSB*2 per year (paid in cash). It is available to families with average monthly income per family member lower than 1.5*SSI.
Figure 3.29. The social assistance benefit is larger for larger families
Maximum social assistance benefit by family type, 2020
Note: Since the OECD TaxBen baseline scenario assumes the first and second child in a family are aged 4 and 6 respectively and pupil assistance is available to students studying from 6 to 18 years, pupil assistance is only available to families with at least two children in the OECD TaxBen model.
Source: OECD Tax and Benefit model TaxBen.
Social assistance benefit policy options
Given the scale of the poverty challenge, Lithuania should continue to increase social benefits through increasing state-supported income. The most effective way to increase SA in Lithuania in 2020 is through increases to the SSI and the income equivalence scale based simulations of 10% increases in different policy levers (from baseline level in 2020) (Figure 3.3). At middle-income levels (50% of AW), SSI remains the most important factor but income disregards becomes relatively more important as net incomes increase and so too does the share that can ‘disregarded’ from the means-test.
Recent increases in social benefit are insufficient to keep pace with rising wages and inflation so regular indexation of SSI increases could be considered. In recent years, Lithuania has introduced reforms to support those most in need including linking the SSI to consumption needs in 2019 and increasing the amount of the SSI since then (most recently to EUR 147 per month in June 2022). While these reforms provide modest support in the direction of containing poverty rates, the amounts are low and the rates of increase may lag those of inflation and wage growth. Lithuania could change its policy of ad-hoc annual SSI increases to regular indexation of SSI increases with inflation.
While the design of social benefit supports larger families, it may be low for single individuals on a relative needs basis. The SA available to childless single individuals could be relatively low compared to larger families given their needs. Based on a square root equivalence scale (where the needs of a family are reflected by the square root of family size), the max SA benefits for larger families are high relative to singles. Compared to a single individual, a family of four receives 3.2 times the SA but has only twice the needs. A relatively higher benefit for larger families was also the case for the housing benefit (see Figure 3.24) highlighting Lithuania’s strong policy focus on protecting families. However, single individuals may be somewhat under compensated by SA based on their needs relative to larger families. One option to reduce this disparity would be to increase the overall level of the SSI but to simultaneously rebalance the equivalence scales by reducing them after the second person thus further protecting high risk of poverty single individuals and single parents. Such a rebalancing of the equivalence scale would be in line with international norms. The income equivalence scale in Lithuania in 2020 (at 100%, 90% and 70% per additional person respectively) is relatively high (the Eurostat scale of poverty calculations which is 100%, 50% and 30%) which implies that Lithuania gives relatively high weights per each additional person in international comparison. In 2021, the amounts of social assistance have since been revised.
Figure 3.30. Increasing state-supported income and the income equivalence scale are effective policy levers for raising social assistance benefits
Simulating social assistance benefit, for a given 10% increase in SSI, income equivalence scale, income disregards and the basic allowance, single family with two children, Lithuania in 2020
Source: OECD Tax and Benefit model TaxBen.
The income disregards rule encourages formal employment, but it could be enhanced for single individuals and couples. The income disregards (ID) rule in Lithuania allows a portion of net in-work income to be ‘disregarded’ for the SA means-test. The portion disregarded is higher for larger families and more so for larger single parent families (for individuals or couples with no children, it is 15%, for couples raising one or two children it is 20%, for couples raising three or more children it is 25%, for individuals raising one or two children it is 30% and for individuals raising three or more children it is 35%) (note that since June 2020, disregards were increased by a further 5 p.p.). The ID rule slows the decline of SA as incomes rise encouraging formal employment. As income rises, the rate at which the SA declines is slower for larger and single parent families (such as a single parent with four children), providing a greater incentive to escape the poverty trap relative to childless single individuals and couples. For a low-income couple with 4 children, if there were no ID rule in place at gross incomes of EUR 300, 500 and EUR 700, the family would receive SA of EUR 390, EUR 282 and EUR 192 in 2020. With the ID rule (where 25% can be disregarded for this family type), the family would instead receive EUR 436 (+12%), EUR 355 (+25%) and EUR 288 (+50%).
The design of child benefits
Lithuania currently has several child benefits focused on supporting low-income and larger families with children. This section focuses on the universal child benefit (UCB) and the additional child benefit (ACB) rather than the COVID-19 supplemental child benefit (CCB) and the maintenance child benefit (MCB). The CCB is temporary (and not part of any baseline child benefit analysis) and the MCB refers to state alimony paid to single parents (and not a government benefit in the normal sense). UCB is EUR 60.06 per child per month in 2020 (1.54 x basic social support or BSB of EUR 39 per month 2020) (Figure 3.31). As the name suggests, the benefit is paid universally to all children (up to 18 years of age and up to 21 years of age if studying) and is not means-tested. ACB is EUR 40.17 per child per month in 2020 (1.03 x BSB). ACB is aimed at supporting low-income families with one to two children, families with three or more children and children with disabilities. Unlike UCB, ACB is means-tested for low-income families with one or two children. The average family’s income per person per month should not exceed EUR 250 per month (or SSI x 2) or EUR 3 000 per year16 (note that a single parent with one or two children implies a family of two or three persons while a couple with one or two children implies a family of three or four persons). For a family of two, three or four persons, the family’s monthly income should not exceed EUR 500 (EUR 6 000pa), EUR 750 (EUR 9 000pa) and EUR 1 000 (EUR 12 000pa) respectively. ACB is not means-tested for larger families with 3 or more children. Lastly, and although not the focus of this review, the MCB is EUR 70.2 per child per month in 2020 (1.8 x BSB). MCB applies to children that do not receive all (or part) of the support awarded from their parents. It is aimed at supporting single individuals raising children (and is not available to couples). A family may be entitled to one, two or all three of the child benefits, in which case the benefits are added together.
Figure 3.31. The main child benefits in Lithuania are the universal and additional child benefits
Source: OECD analysis.
While the universal child benefit remains in place regardless of income, the mean-tested additional child benefit is withdrawn at higher incomes for smaller families. The UCB remains constant across the income distribution (as it is paid regardless of income) while the ACB is withdrawn at higher incomes (as it is means-tested based on average family net income) (Figure 3.33).17 For a single parent with one child, the ACB is withdrawn at 52% of the AW (as net income exceeds EUR 500).
Figure 3.32. Child benefits fall as incomes rise due to the withdrawal of the mean-tested additional child benefit
Child benefits in Lithuania in 2020, single person with one child, by gross income
Note: Total child benefit (CB) is the sum of the universal child benefit and the additional child benefit. The child benefit plus a COVID-19 supplement is the sum of the increased UCB and ACB. The child maintenance benefit includes child benefit but not the COVID-19 supplement.
Source: OECD analysis of Taxing Wages 2021.
Child benefit policy in Lithuania favours larger families and parents in a couple
Child benefit amounts increase linearly for each additional child at lower income levels. For low-income families, the monthly child benefit (CB) amount increases by EUR 100 per additional child. Below the ACB income threshold (45% of the AW), a single individual or a couple with one, two, three and four children would each receive a total monthly CB of EUR 100, EUR 200, EUR 300 and EUR 400 respectively. CB (i.e. UCB and ACB) declines at higher income levels as the means-tested ACB is withdrawn (Figure 3.33).18 For single parents with one and two children, the ACB is withdrawn fully and at-once at ~50% and ~82% of AW respectively.
High income families with two children face a higher tax burden than an equivalent family with a third child because the ACB is not withdrawn. In addition to larger families receiving greater child benefits with more children, middle and higher-income larger families with three or more children benefit from the non-withdrawal of the ACB as incomes rise (relative to families with two or less children) (Figure 3.33).
Child benefit policy is more favourable to couples than single parents at middle-incomes. Couples with one child benefit from the ACB at higher income levels (up to ~82% of AW) relative to single individuals with one child (only up to ~50% of AW) as the benefit is based on the average family’s income per person. Similarly, couples with two children continue to benefit from the ACB up to ~113% of AW compared to just ~82% for single parents with two children. This occurs partly because there are two adults instead of one in the couple with the same gross earnings so they are eligible for means-tested benefit for longer.
Figure 3.33. Single families with more children receive greater child benefits
Monthly child benefits (EUR) by gross wage as a % of average wage, single families with children, 2020
Source: OECD analysis of Taxing Wages 2021.
Figure 3.34. Child benefit policy modestly favours middle-income couple parents over single parents
Child benefits for single and couple parents with one and two children in Lithuania in 2020, by gross income
Source: OECD analysis of Taxing Wages 2021.
The withdrawal of the additional child benefit is not insignificant in absolute and relative amounts. For a single parent with one or two children, the withdrawal of the ACB would represent a nominal decline in the monthly child benefit of 40% (from EUR 100 to EUR 60 and from EUR 200 to EUR 120 respectively). In the latter case of a single parent with two children the loss of EUR 80 per month is not insignificant at 5.7% of the AW. The withdrawal of the child benefit could produce some risks of poverty trap, which are explored later (see Figure 4.27) and a gradual tapering could be considered if administratively feasible.
A temporary additional child benefit was introduced in response to the pandemic
A once-off additional lump-sum child benefit was introduced in June 2020 to reduce the effects of COVID-19 on families. The maximum child benefit per child per month in 2020 is EUR 100 (60 UCB + 40 ACB) and EUR 117 when the additional ‘covid’ lump-sum is included (70 + 47).19 These combined child benefits represent 7.1% and 8.3% of the AW respectively. The once-off lump-sum child benefit introduced in June 2020 to reduce the effects of the COVID-19 pandemic increased the child benefit by 17%.
The additional covid child benefit raises the benefit by the same proportion regardless of family type. For single individuals and couples, the COVID-19 supplement increases the benefit by 17% regardless of the number of children. Although the CCB and MCB are not included in child benefits as part of the baseline analysis in this report, if they were to be included, a single family with one child can receive a max total child benefit of EUR 170 (comprised of UCB EUR 60, ABC EUR 40, CCB EUR 17 and MCB EUR 70) and a single family with two children can receive a maximum of double or EUR 340 (EUR 120, EUR 80, EUR 34, EUR 140).
Although not a standard child benefit paid by government, the maintenance child benefit is a significant support for single families. The MCB is targeted at the children of single individuals. The MCB is paid to children not receiving all or part of the support awarded by the child’s parents. The MCB is a max of EUR 70 (BSB x 1.8) per month per child and increases linearly with more children in absolute terms. For low earners, compared to the UCB and ACB (EUR 60 + EUR 40), the MCB (EUR 70) represents a significant 70% increase in combined child benefit. Given that low-income couples and single individuals have the same child benefits regardless of the number of children, the MCB also represents a 70% increase in benefits compared to couple families. For higher income wage earners (above the ACB income threshold), compared to only the UCB (EUR 60), the MCB (EUR 70) represents a larger 117% increase. Despite the increase represented by the MCB, it is not means-tested and remains constant regardless of income.
References
[9] Adam, S., D. Phillips and B. Roantree (2019), “35 years of reforms: A panel analysis of the incidence of, and employee and employer responses to, social security contributions in the UK”, Journal of Public Economics, Vol. 171, pp. 29-50, https://doi.org/10.1016/J.JPUBECO.2018.05.010.
[7] Anderson, P. and B. Meyer (1997), “Unemployment Insurance Takeup Rates and the After-Tax Value of Benefits”, The Quarterly Journal of Economics, Vol. 112/3, pp. 913-937, https://doi.org/10.1162/003355397555389.
[5] Arpaia, A. and G. Carone (2004), “Do labour taxes (and their composition) affect wages in the short and the long run? - Alfonso Arpaia and Giuseppe Carone”, European Economy - Economic Papers 2008 - 2015 103, pp. 223-244, https://doi.org/10.3280/QUA2015-103010.
[15] Bargain, O. and K. Orsini (2004), “In-Work Policies in Europe: Killing Two Birds with One Stone?”.
[33] Bielskis, K. and A. Ciginas (2020), “Household Wealth and Finances. Results for Households in Lithuania for 2017”, http://www.lb.lt (accessed on 8 February 2022).
[14] Blundell, R. and A. Shephard (2012), “Employment, Hours of Work and the Optimal Taxation of Low-Income Families”, The Review of Economic Studies, Vol. 79/2, pp. 481-510, https://doi.org/10.1093/RESTUD/RDR034.
[10] Bozio, A., T. Breda and J. Grenet (2019), “Does Tax-Benefit Linkage Matter for the Incidence of Social Security Contributions?”, IZA Discussion Papers, https://ideas.repec.org/p/iza/izadps/dp12502.html (accessed on 2 December 2022).
[8] Bozio, A., T. Breda and J. Grenet (2017), “Incidence and Behavioural Response to Social Security Contributions: An Analysis of Kink Points in France”, Economist (Netherlands), Vol. 165/2, pp. 141-163, https://doi.org/10.1007/S10645-017-9297-4.
[13] Chetty, R., J. Friedman and E. Saez (2013), “Using Differences in Knowledge across Neighborhoods to Uncover the Impacts of the EITC on Earnings”, American Economic Review, Vol. 103/7, pp. 2683-2721, https://doi.org/10.1257/AER.103.7.2683.
[18] Eissa, N. and H. Williamson Hoynes (2004), “Taxes and the labor market participation of married couples: the earned income tax credit”, https://doi.org/10.1016/j.jpubeco.2003.09.005.
[1] European Commission (2019), Labour Taxation in Romania: Revised, but not changed, https://ec.europa.eu/info/sites/default/files/economy-finance/eb050_en.pdf (accessed on 27 April 2022).
[6] Gruber, J. (1995), “The Incidence of Payroll Taxation: Evidence from Chile”, https://doi.org/10.3386/W5053.
[2] Ivaškaitė-Tamošiūnė, V. et al. (2018), “The Effect of Taxes and Benefits Reforms on Poverty and Inequality in Latvia”, https://doi.org/10.2765/830917.
[30] Koutsogeorgopoulou, V. (2020), “Reducing poverty and social disparities in Lithuania”, OECD Economics Department Working Papers, No. 1649, OECD Publishing, Paris, https://doi.org/10.1787/b631de7d-en.
[17] Liebman, J. (1998), “The Impact of the Earned Income Tax Credit on Incentives and Income Distribution”, NBER Chapters, pp. 83-120, https://ideas.repec.org/h/nbr/nberch/10914.html (accessed on 5 December 2022).
[29] Navicke, J., S. Avram and L. Demmou (2016), “The effects of reform scenarios for unemployment benefits and social assistance on financial incentives to work and poverty in Lithuania”, OECD Economics Department Working Papers, No. 1310, OECD Publishing, Paris, https://doi.org/10.1787/5jlv2jmtmsmr-en.
[34] OECD (2021), Affordable Housing Database - OECD, http://www.oecd.org/social/affordable-housing-database.htm.
[23] OECD (2021), Health at a Glance 2021: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ae3016b9-en.
[21] OECD (2021), Pensions at a Glance 2021: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ca401ebd-en.
[4] OECD (2021), Revenue Statistics 2021: The Initial Impact of COVID-19 on OECD Tax Revenues, OECD Publishing, Paris, https://doi.org/10.1787/6e87f932-en.
[22] OECD (2021), Taxing Wages 2021, OECD Publishing, Paris, https://doi.org/10.1787/83a87978-en.
[16] OECD (2020), OECD Economic Surveys: Israel 2020, OECD Publishing, Paris, https://doi.org/10.1787/d6a7d907-en.
[31] OECD (2020), OECD Economic Surveys: Lithuania 2020, OECD Publishing, Paris, https://doi.org/10.1787/62663b1d-en.
[26] OECD (2020), “The design of unemployment benefits schedules over the unemployment spell: The case of Belgium”, http://www.oecd.org/employment/jobs-strategy. (accessed on 5 February 2022).
[32] OECD (2018), OECD Economic Surveys: Lithuania 2018, OECD Publishing, Paris, https://doi.org/10.1787/eco_surveys-ltu-2018-en.
[27] OECD (2018), OECD Reviews of Labour Market and Social Policies: Lithuania, OECD Reviews of Labour Market and Social Policies, OECD Publishing, Paris, https://doi.org/10.1787/9789264189935-en.
[25] OECD (2018), Reshaping the Personal Income Tax in Slovenia, OECD, https://www.oecd.org/tax/tax-policy/reshaping-the-personal-income-tax-in-Slovenia.pdf.
[24] OECD (2017), OECD Economic Surveys: Latvia 2017, OECD Publishing, Paris, https://doi.org/10.1787/eco_surveys-lva-2017-en.
[12] OECD (2011), Taxation and Employment, OECD Tax Policy Studies, No. 21, OECD Publishing, Paris, https://doi.org/10.1787/9789264120808-en.
[11] OECD (2005), OECD Employment Outlook 2005, OECD Publishing, Paris, https://doi.org/10.1787/empl_outlook-2005-en.
[19] Saez, E. (2002), “Optimal Income Transfer Programs: Intensive versus Extensive Labor Supply Responses”, Source: The Quarterly Journal of Economics, Vol. 117/3, pp. 1039-1073, https://about.jstor.org/terms (accessed on 14 July 2022).
[3] Shum, C., J. Fay and M. Lui (2017), “Worldwide or territorial approach for individual income tax : which is more prevalent?”, Journal of Modern Accounting and Auditing, Vol. 13/4, p. 137, https://doi.org/10.17265/1548-6583/2017.04.001.
[28] Užgalė, U. et al. (2020), The OECD Tax-Benefit Model for Lithuania: Description of Policy Rules for 2020, OECD, https://www.oecd.org/els/soc/TaxBEN-Lithuania-2020.pdf (accessed on 5 February 2022).
[20] Vandelannoote, D. and G. Verbist (2020), “The impact of in-work benefits on work incentives and poverty in four European countries:”, https://doi.org/10.1177/0958928719891314, Vol. 30/2, pp. 144-157, https://doi.org/10.1177/0958928719891314.
Notes
← 1. Note that the microdata sample size is insufficiently large to examine the tax shares at the income level where the top PIT rate and SSC ceiling apply.
← 2. The monthly average wage of EUR 1 400 is based on OECD secretariat estimate for 2020. In Lithuania, the average wage applied to calculate the state social insurance contribution base is approved by the law of Approval on Budget Indicators of the State Social Insurance Fund for the relevant year. It is the average gross monthly earnings (including salary data for the sole proprietorships) published by the Statistics of Lithuania of Q3 and Q4 for the year before the previous year and Q1 and Q2 for the previous year. In the third quarter of 2019, the average gross monthly earnings in the whole economy (individual enterprises included) totalled EUR 1 306.30. In the third quarter of 2020, the average gross monthly earnings in the whole economy (individual enterprises included) totalled EUR 1 443.80.
← 3. The monthly average wage of EUR 1 400 is based on OECD secretariat estimate for 2020.
← 4. Note that the average wage (AW) calculated by Statistics Lithuania and OECD Taxing Wages are similar but not the same.
← 5. The basic allowance is applied to employment and related income.
← 6. Individuals or groups of the population obtain health care through a variety of financing arrangements. These involve a range of third-party schemes but also, by convention, payments made directly by households. Government financing schemes, on a national or subnational basis or for specific population groups, entitle individuals to health care based on residency, and form the principal mechanism to cover health care costs in close to half of OECD countries. The other main method of financing is some form of compulsory health insurance (managed through public or private entities) (OECD, 2021[23]).
← 7. EUR 607 * 23.27% = EUR 141.25 in 2020 and EUR 642* 23.27% = EUR 149.39 in 2021.
← 8. A gross monthly AW of EUR 1,306.3 in Q3 2019 implies unemployment benefit cannot exceed EUR 760.01 in January 2020 and in Q3 2020, the gross monthly AW is EUR 1 443.8 and so the maximum unemployment benefit cannot exceed EUR 840.
← 9. Insured income is all incomes of an individual from which the unemployment insurance contributions were paid or had to be paid. The average monthly insured income is calculated taking account of the amount of the actual insured income of an unemployed person for every month during the previous 30 months.
← 10. Only 10.5% of wealth corresponds to self-employed business and 3.7% is accumulated through financial vehicles.
← 11. According to OECD TaxBen, the net annual income of a person without a family is 62 SSI, net annual income for a family of two or three persons is 122 SSI and net annual income per person for a family of four or more persons is 35 SSI. The model also assumes the location is Vilnius.
← 12. Since 30 September 2016, the list of circumstances under which cash social assistance cannot be reduced was expanded. The scheme of proportionate reduction of social benefit is not applied if at least one of the following conditions holds: the Employment Service did not offer a job or an opportunity to participate in active labour market policy measures; a working-age unemployed person (who is employable and not in education) participated in useful social activity organized by the municipal administration. In the model, it is assumed that one of the above conditions are satisfied, thus the reduction in social benefits is not modelled.
← 13. The average monthly income of the resident is determined based on their income in the previous 3 months from which the social benefit is granted.
← 14. More specifically, for persons living together, for the first person, SA is 100% of the difference between the SSI per person and the average monthly income of the persons living together per person. For the second and third person, the share is reduced to 80% and 70% respectively.
← 15. It is a maximum because it assumes persons with more income can earn a smaller fraction of the SA.
← 16. Income refers to income in the previous calendar year. In the TaxBen model, this is assumed to be current gross income.
← 17. Similarly, the child maintenance benefit is paid regardless of income and as long as the requirements are met – the child is part of a single family (i.e. children who do not receive all or part of the support awarded by the child’s parent).
← 18. With the exception of the child maintenance benefit, which is only available to single families and not couples.
← 19. Not including the maintenance child benefit.