Labour supply and inequality effects of in-work benefits: Evidence from Serbia

Ranđelović, Saša; Žarković Rakić, Jelena; Vladisavljević, Marko; Vujić, Sunčica Article Labour supply and inequality effects of in-work benefits: Evidence from Serbia Naše gospodarstvo / Our Economy Provided in Cooperation with: Faculty of Economics and Business, University of Maribor Suggested Citation: Ranđelović, Saša; Žarković Rakić, Jelena; Vladisavljević, Marko; Vujić, Sunčica (2019) : Labour supply and inequality effects of in-work benefits: Evidence from Serbia, Naše gospodarstvo / Our Economy, ISSN 2385-8052, Sciendo, Warsaw, Vol. 65, Iss. 3, pp. 1-22, https://doi.org/10.2478/ngoe-2019-0010 This Version is available at: https://hdl.handle.net/10419/290445 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ 1 NAŠE GOSPODARSTVO OUR ECONOMY pp. 1–22 ORIGINAL SCIENTIFIC PAPER Citation: Ranđelović, S., Žarković Rakić, J., Vladisavljević, M., & Vujić S. (2019). Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia. Naše gospodarstvo/Our Economy, 65(3), 1–22. DOI: 10.2478/ngoe-2019-0010 DOI: 10.2478/ngoe-2019-0010 UDK: 004.775:339.137.2:005.412 JEL: J22, H24, D31, I32 RECEIVED: JULY 2019 REVISED: AUGUST 2019 ACCEPTED: SEPTEMBER 2019 Vol. 65 No. 3 2019 Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia Saša Ranđelović University of Belgrade, Faculty of Economics, Serbia [email protected] Jelena Žarković Rakić University of Belgrade, Faculty of Economics, Serbia [email protected] Marko Vladisavljević Institute of Economic Sciences, Belgrade, Serbia marko.vladisavlje[email protected] Sunčica Vujić1 University of Antwerp and University of Bath, Department of Economics, Belgium suncica.[email protected] Abstract Low labour market participation, together with the high effective tax wedge at low wage levels, create a fertile ground for the introduction of the in-work benefits (IWB) in Serbia. Our paper provides an ex-ante evaluation of the two IWB schemes, directed at stimulating the labour supply and more equal income distribution. The methodological approach combines the tax-and-benefit microsimulation model with the discrete labour supply model. Our results show that both individual and family-based IWB schemes would considerably boost labour market participation, although family-based benefits would have disincentivizing effects for the secondary earners in couples. Most of the behavioural changes take place among the poorest individuals, with significant redistributive effects. Keywords: in-work benefits, labour supply, inequality, discrete choice model, microsimulation Introduction Low labour market participation, high informality, particularly at the bottom end of the earning distribution, and high formalization costs constitute theoretically a solid case for the introduction of making-work-pay policies, such as in-work benefits (IWB), which are means-tested benefits conditional on employment status. These policies have become popular in many European countries trying 1 Corresponding author: Sunčica Vujić, Department of Economics, University of Antwerp, Prinsstraat 13, 2000 Antwerp, Belgium. Email: [email protected]. Tel.: +32 32 654 902. 2 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 to promote labour market participation and work formalization, particularly among low-paid workers. Empirical studies confirm that the IWB policies have been effective in tackling the above-mentioned labour market problems in developed countries, particularly in the United States (US) and the United Kingdom (UK) (Blundell et al., 2000; Blundell & Hoynes, 2004; Meyer & Rosenbaum, 2001; Orsini, 2006). However, the empirical literature on the effectiveness of these policies in the European developing and transition economies is scarce. Our paper attempts to fill this gap. A high labour force participation rate is important for economic growth, competitiveness, poverty reduction, and the political and social stability of a country, especially with an aging population. These are some of the reasons why the European Union 2020 (EU 2020) strategy set the employment rate target for 20- to 64-year-olds at 75% (European Commission, 2010). According to the Statistical Office of the Republic of Serbia (SORS, 2016), the country’s average (over four quarters) labour force participation rate of the working age population (15- to 64-year- olds) was 60.5% in 2016, while the employment rate was 45.2%; both of these indicators are far below the EU-28 average of 73% for the labour force participation rate, and 71% for the employment rate in 2016 (Eurostat, 2019a; OECD, 2019). Inactivity is particularly high among lowskilled, low-income earners. Serbia’s active population will further decline in the current decade, due to the exit of baby boomers from the labour market, and the increasing outbound migration (Arandarenko et al., 2012). Since the onset of the global economic crisis, in addition to the declining labour market participation, the informality rate has been on the rise, being particularly high in the agriculture and self-employment sectors, thus further reducing the reservoir of formal labour market participation. Following the well-established theoretical base and empirical framework for the evaluation of labour market effectiveness of the IWB policies (Bargain & Orsini, 2006; Blundell, 2000; Immervoll & Pearson, 2009; Saez, 2002), our research combines the EUROMOD-based tax-and-ben- efit microsimulation model for Serbia (SRMOD) with a structural labour supply model, in order to evaluate the effects of the IWB policies, as well as the interplay between the IWB policies and other tax-and-benefit policies, on labour supply and income distribution in Serbia. The aim of this paper is to provide an empirical estimation of labour supply effects of in-work benefits for a transition country. These results will be benchmarked to the results on the effects of IWBs for countries that are similar to Serbia in terms of labour market performance (high unemployment and inactivity rates, particularly for young people and women) and features of tax and benefit systems (lacking social assistance benefits). The results will be compared with those for other Western Balkan countries, as well as Portugal, Spain, Italy and Greece (which have similar labour market characteristics as Serbia). In addition, the results on labour supply elasticities will be compared to those for developed and other transition countries. We see a threefold contribution of our paper to the existing literature. First, bearing in mind that the effectiveness of the IWB policies depends on their design, as well as on the institutional characteristics of a targeted country, our paper constitutes a new methodological framework taking into account the characteristics of the labour market of a transition country. The existing literature has mostly focused on the developed economies, while the empirical literature for the European transition economies, to the best of our knowledge, is limited to the evidence for Slovenia (Kosi & Bojnec, 2009), Poland (Myck et al., 2013), Macedonia (Mojsoska et al., 2015) and Serbia (Clavet et al., 2019; Ranđelović & Žarković-Rakić, 2013; Žarković-Rakić et al., 2016). Second, since low labour market participation, high informality and high formal activation costs are common features of the Western Balkan economies, while the IWB policies are almost non-existent, the empirical results for Serbia may represent a considerable contribution not only to the existing empirical literature but also to a discussion on introducing making-work-pay policies in the region. Third, we provide evidence on the effectiveness of the IWB policies with respect to different family structures (singles and couples) and with respect to the position of women in the labour market. This is particularly important taking into account that inactivity rates of women in Serbia are considerably higher than those of men (76.6% and 63.8%, respectively, in the last quarter of 2016; SORS, 2016). The results obtained in this paper suggest that both individual IWB (IIWB) and family IWB (FIWB) would trigger a decline in labour market non-participation: the effects of FIWB are larger for singles, while the IIWB would have higher impact on the labour supply of individuals in couples. At the same time, the FIWB would have somewhat larger effects for single women than for men, the effects of IIWB being the opposite, while no significant difference in terms of labour supply reaction to the IIWB and the FIWB by gender is found in the case of coupled individuals. The policy is expected to yield positive effects on inequality, since most of the labour activation would happen at the lower end of the income distribution. The difference in the size of the effects of IIWB and FIWB, depending on the income level and marital status, is the consequence of the difference in the design of the IIWB and FIWB policies but also the result of variation in labour supply elasticities by income levels and marital status. 3 The remainder of the paper is organised as follows. The next section provides an explanation of the participation and formalization disincentives coming from the tax-and-benefit system design. Section 3 deals with the analysis of the IWB design and the overview of empirical literature on the IWB policies. In section 4, the data and methodology are presented, while the results are presented and discussed in section 5. Section 6 concludes the paper. Tax-and-benefit System in Serbia: Why Working at Low Wage Levels Does Not Pay? In this section, we provide further information on the characteristics of the labour market in Serbia in general and the tax-and-benefit structures in particular. Serbia’s labour market performs considerably worse when compared to most other European economies. The labour market participation rate currently stands at 63.4%, while the unemployment rate is 18.8%. Inactivity and unemployment rates are particularly high among low-educated individuals. Living Standards Measurement Survey (LSMS) 2007 data, used in this paper, show that non-participation rates for those with primary education (55.5%) are significantly higher than for those with secondary (30.0%) and tertiary education (18.7%). At the same time, women are in a particularly difficult labour market position. On average, they face 16.6 percentage point higher inactivity rates and 4.7 percentage point higher unemployment rates than men (Table 1). Arandarenko and Vukojević (2008) show that the tax-ben- efit wedge in Serbia is rather regressive at the bottom of wage distribution, due to the high minimum social security contribution base, relatively high social contribution rates, low non-taxable threshold (for personal income taxation) and withdrawal of means-tested benefits. This has been confirmed by the World Bank study for Serbia (Koettl, 2010), which calculated the Implicit Costs of Formalization (ICF), defined as the share of income that an informal worker has to give up in order to formalize work, showing disincentives for formalization stemming from labour taxation and benefit withdrawal. The study shows that a single person with no children who earns less than the minimum wage in the informal sector has to give up between 40% and 75% of income in order to formalize work. Labour market participation in Serbia is among the lowest in Central and Eastern Europe (CEE), being close to the activity rates in the Western Balkan countries, while the female participation rate is considerably lower than the male participation rate (Table 2). At the same time, income inequality in Serbia is among the highest in Europe (Davies, 2018). Table 2. Labour market participation rates in the CEE (2014)  Total Women Men Czech Republic 73.3 65.6 80.9 EU-28 72.1 66.0 78.2 Slovenia 70.9 67.4 74.2 Slovakia 70.2 62.7 77.7 Bulgaria 68.1 64.2 72.0 CEE-average 67.8 61.3 74.3 Poland 67.3 60.5 74.2 Romania 65.0 57.3 72.8 FYR of Macedonia 64.6 51.5 77.3 Hungary 64.6 58.6 70.6 Croatia 64.1 58.6 69.7 Albania 63.0 51.7 74.2 Serbia 61.7 53.9 69.7 Montenegro 59.2 52.3 66.1 Bosnia and Herzegovina 54.5 42.1 67.1 Source: World Bank Database Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia Table 1. Non-participation and unemployment rates by level of education and gender (%)  Total Men Women Education level Primary and less Secondary Higher Unemployment rate 13.9 11.8 16.5 14.9 15.4 7.9 Inactivity rate 36.6 27.3 43.9 55.5 30 18.7 Notes: Working age population (15-64 years) Source: Authors’ calculations based on LSMS, 2007. Low-education attainment coupled with a lack of work experience generates low earning capacity in the labour market. When earnings or potential earnings are low, incentives to seek employment or stay in employment are usually limited. Incentive problems are aggravated by high tax burdens on labour income and by cuts in social benefits designed to provide at least some safety nets for those with no or very low income (Immervoll & Pearson, 2009). 4 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 The labour tax wedge in Serbia for those with low earnings (67% of average wage) is among the largest in the CEE and is also higher than the EU-28 average (Figure 1). In addition, the degree of progressivity of labour taxation in Serbia is among the lowest in the CEE. When the wage increases from 67% of the average wage to 167% of the average wage, the labour tax wedge (as per cent of labour costs) in Serbia rises only by 1.3 percentage point, while in the CEE and the EU-28 the rise is considerably steeper — 6.4 percentage point and 8.1 percentage point, respectively (Figure 1). The relatively high labour tax burden of low-paid jobs and low progressivity are the consequences of several factors, the most important being the mandatory minimum social security base and the sudden withdrawal of means-tested benefits upon formal employment.2 The minimum base, which is mandatory for every worker, is set at 35% of the average wage, implying that when the actual wage is below the minimum base, the social contributions are calculated on the minimum base. Given that the base is effectively not adjusted for working hours, the low-paid part-time jobs are exposed to a high social contribution burden. 2 This is a peculiar feature of the social security contribution systems in the Western Balkan region. The most drastic example is Macedonia, where the mandatory base is set as high as 50% of average wage. Additionally, the labour tax reform that was introduced in 2001 brought about the abolishment of fringe benefits. The two most important benefits of this kind were food allowances (paid monthly) and an annual leave benefit. Given that both fringe benefits were not taxed and were paid in equal amounts to each worker, the abolishment of these benefits contributed to the regressive character of the labour tax system, which was in effect until 2007 (Arandarenko & Vukojević, 2008). The changes to the labour tax system that took place in 2007 envisaged a cut in the wage tax rate between 12% and 14% and the introduction of the zero tax bracket (up to 5,000 Dinars (i.e., 63.1 Euro), or approximately 15% of the average wage). However, the tax burden on labour did not change considerably, given that the social security contributions dominated the tax wedge.3 Another peculiarity of the Serbian labour market relates to relatively high informality. As put forward by Krstić and Sanfey (2011), between 2002 and 2007 informal work rates in Serbia rose despite strong economic growth and the improved business climate in the country. The authors argue that one possible reason for this unexpected result 3 In 2001, contributions were set at 32.6% of the gross wage, equally split between employers and workers. The first increase in mandated contributions occurred in 2003, with an increase of 1 percentage point. The next modification was made in 2004, and currently the overall social security tax rate amounts to 35.8% of gross wage: 22% for old age, disability and survivors’ pensions, 12.3% for health insurance, and 1.5% for unemployment insurance. Figure 1. Labour tax wedge and progressivity of labour taxation in the CEE (2014) Source: Authors’ calculations based on the Eurostat Labour Force Survey (LFS) 2014 data and International Labour Organization (ILO) 2014 data 5 is the regressive character of the labour tax system that was introduced in 2001 and that was applied until January 2007. The incentives to join the formal economy were diminished for both workers and employers. However, it should be noted that the high informality rate in Serbia was to a large extent driven by informal work in agriculture and self-employment, where informal workers accounted for 87% and 53% of the total number of workers, respectively, while the average informal wage-employment was considerably lower, with a 10% share in the total number of wage-employees in 2007, thus being below the average for Southern Europe4 (Hazans, 2011). Besides labour taxation, the social benefits design is another piece of the puzzle necessary to explain the high levels of inactivity and informality among the working-age population in Serbia. Once a person has a formal income on her/ his record, major income-tested benefits (social assistance and child allowance in particular) will be decreased by the total amount of earned income or completely withdrawn. In their study on inactivity in the Serbian labour market, Arandarenko et al. (2012) show that a person receiving 4 Southern Europe constitutes Italy, Spain, Greece, Portugal, and Cyprus. social benefits does not have an incentive to search for a job offering a salary below 20% of the average gross wage, which is equivalent to a part-time job equalling 20 hours at the minimum hourly wage. Mainly due to the mandatory minimum social security contribution base, net income for this individual becomes equal to the amount of social assistance benefit. Therefore, the so-called mini-jobs and midi-jobs (mainly part-time jobs) are not economically attractive for low-wage earners. As a result of the minimum social contribution base, sudden withdrawal of the means-tested benefits and other mentioned factors, the tax wedge distribution is regressive up to 33% of average wage, afterwards being only slightly progressive (Figure 2). At the same time, the implicit formalization costs are the highest at the very bottom of wage distribution.5 Such design of the tax-benefit scheme creates considerable disincentives for labour market participation of low-skilled individuals. 5 Implicit formalization costs are calculated as the percentage of initial income that a household has to give up in order to switch from the informal to the formal sector, assuming the gross wage offer is unchanged. Figure 2. The tax wedge, implicit costs of formalization and the marginal effective tax rate for singles with no children in Serbia (2009) Source: Koettl (2010) Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia 6 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 In-Work Benefits Objective, Design and Labour Supply Effects: Literature Review The intention of policymakers to address the issue of insufficient labour market participation and very low participation of low-skilled, low-paid workers (thus, at the same, time tackling inequality and poverty issues) has brought the design of tax-and-benefit policies and their interplay back into the focus of empirical literature and discussions. Measures directed at increasing the income of persons with a low earnings capacity have mostly centred on the introduction of the IWB policies. The IWB policies are designed to promote work and reduce poverty by generating a difference between the incomes of working people and the counterfactual situation, that is, the incomes they would have if they were out of work. Additionally, the IWB schemes contribute to higher formality by effectively reducing the labour tax wedge and thus encouraging wage formalization. The introduction and expansion of the IWBs in the European countries has been inspired by the Earned Income Tax Credit (EITC), introduced first in the US, and the subsequent Working Family Tax Credit (WFTC) practice in the UK. The main motivations for the introduction of these policies in Europe and North America during the early 1990s were the low levels of labour force participation and employment experienced by certain specific demographic groups of working age (Blundell, 2006). In a theoretical framework, Kolm & Tonin (2011) show that IWBs can be extended to larger sections of the workforce and to other countries. Nowadays, most of the OECD countries apply some types of the IWB programs in order to boost labour market activity and employment (Immervoll, 2012; Immervoll & Pearson, 2009). Even though there are differences among the countries in terms of the design of the IWBs, all employment-con- ditional measures use at least one of the following criteria to assess eligibility and determine the amount of benefits: having children, working a minimum number of hours, and receiving income from work or entering into employment. Most of these benefits are proportional to gross income up to a maximum amount, afterwards being gradually withdrawn. In other words, they are characterised by the gradual phase-in and phase-out brackets, as a mean of targeting individuals with specific earnings levels or working hours. An important aspect of the IWB design is the choice of the unit used to assess income (Orsini, 2006; Orsini & Bargain, 2006). In some countries, eligibility for benefits is assessed at the household level, while in other countries it is focused on individuals. Family IWBs are more effective in boosting the labour supply of single individuals, due to discouraging effects on the second earners in households, who in most cases are women (Eissa & Hoynes, 1998). However, in certain cases, family benefits can have both redistributive and incentive effects. This is the case with lone parents that constitute a large part of poor households (Orsini & Bargain, 2006). On the other hand, individual-level benefits have greater work-incentive effects than do family-based benefits, since they do not discourage the participation of second earners in a couple (Blundell et al., 2000; Orsini & Bargain, 2006). Most evaluations of labour supply effects and distributional effects of the IWB policies in European countries are ex ante evaluations based on a behavioural microsimulation framework. For example, Blundell et al. (2000) have estimated the labour supply preferences on data not affected by the policy reforms, which were then used to simulate the impact of the introduction of the WFTC in the UK. The authors have found that the introduction of the WFTC leads to an increase in labour market participation rates for lone mothers and a small decline in labour market participation amongst women in couples, with no net effect on the labour market participation rates of men in couples.6 The results are consistent with the findings of Brewer et al. (2006). Bell (2005) has found that the decline in child poverty between the fiscal years 2002 and 2004 can be attributed to the introduction of the WFTC program in the UK. Brewer (2006) also noted that the WFTC program reduced the number of families in poverty. St Martin & Whiteford (2003) have estimated that the WFTC program produced about 100,000 new jobs, while the cost of this policy was about 1% of GDP. Orsini & Bargain (2006) have simulated the British WFTC scheme and the individualized wage subsidy scheme for Finland, France, and Germany, countries which have experienced severe poverty traps. They have found that the participation of married women declined in all three countries after the introduction of the WFTC, the negative effect being only partially offset by the positive impact of the reform on single women’s labour supply (in Finland and in Germany). On the other hand, they have found that individual IWB encouraged married women to take jobs. The effects in Finland were lower than in other countries, mainly due to the relatively small labour supply elasticity. Both programs were found to have had significant anti-poverty effects. Haan & Myck (2007) also find strong disincentivizing effects of the British style IWB on coupled individuals, if implemented in Germany. Myck et al. (2013) have found evidence on the IWB-related disincentives on the work of secondary earners in Poland and provide proposals to tackle this issue. Saez (2002) has evaluated the making-work-pay policies in the US and showed that the IWB policies provide an optimal income transfer program when the labour supply choice is 6 For the evaluation of the EITC, see Scholtz (1994, 1996), Eissa and Hoynes (1998), Eissa and Liebman (1996), and Meyer and Rosenbaum (2000). 7 whether or not to work (extensive margin). On the other hand, if the labour supply choice is about intensity of work on the job (intensive margin), then the optimal transfer program is a classical Negative Income Tax program with a large guaranteed income-support schemes, which are taxed away at high rates. The effectiveness of making-work-pay policies is fundamentally dependent on the labour market structure, inherent elasticities and the institutional set-up. Therefore, switching to an environment where participation is low, unemployment is high and institutions are weaker may change the expected outcomes of the IWB policies. With respect to this, several papers have emerged focusing on the southern European countries (particularly Italy), which share these features. Figari (2015) finds that the family IWB in Italy triggers an average increase of female labour supply by 3 percentage points, the individual IWB having stronger incentive effects on coupled women, since their labour supply increase is estimated at 5 percentage points. He also finds that most of the labour supply reactions induced by the IWB take place among the poorest individuals, with important redistributive effects. Similar results for Italy, especially for couples with children, are found in De Luca et al. (2013). Colonna & Marcassa (2013) show that the working tax credit in Italy boosts the labour force participation rate, particularly among unskilled and low-educated women. Late transition economies of Southeastern Europe have even more unfavourable labour market features than the Southern European countries, while making-work-pay policies (and empirical literature on this topic) are largely absent. Mojsoska et al. (2015) use a microsimulation framework to assess the impact of the hypothetical IWB schemes in Macedonia and find that family IWBs are more effective in promoting the labour activation of singles, while individual benefits are more effective in the case of couples, with the effects in both cases being concentrated at the bottom of the income distribution with poverty reduction effects. Using the tax and benefit micro-simulation model for Serbia (SRMOD), Ranđelović and Žarković-Rakić (2013) provide empirical evidence on the incentive and distributional effects of the abolishment of the mandatory minimum social security contribution (SSC) base, showing that the reform would reduce effective average tax rates by more than it would reduce effective marginal tax rate, while the impact of the reform on the overall level of inequality, measured by the Gini coefficient, would be small. Žarković-Rakić et al. (2016) evaluates the impact of the minimum SSC base reform scenarios in Serbia on labour supply and employment formalization and conclude that the proposed reform would not significantly contribute to the transformation of informal full-time to formal full-time jobs. Clavet et al. (2019) evaluate the labour supply and distributive effects of several reform strategies concerning two major social transfers in Serbia: child allowance and social monetary assistance. The results show that, in a context of a low labour participation rate, and high unemployment and informality rates, a benefit strategy is by far the more cost-effective option for reducing child poverty than an employment strategy that aims to raise the work incentives for parents. Methodology: Behavioural Microsimulation Model, Data and Policy Reform Design Model and Data In order to analyse the potential effects of policy measures on labour supply incentives and income redistribution, this paper combines the tax-and-benefit microsimulation model for Serbia (SRMOD), which is based on the EUROMOD platform (Sutherland & Figari, 2013), with a structural, discrete choice, labour supply model.7 Similar to other microsimulation models, SRMOD is a tax-and-benefit calculator based on the micro-data on income, earnings, labour force participation and socio-demographic variables, allowing us to reproduce the budget constraint for each household (i.e., the latent set of working hours and household disposable income alternatives), while the labour supply model rationalizes observed behaviour. The policies simulated in SRMOD refer to Living Standards Measurement Survey (LSMS) dataset in 2007 as the baseline year. The LSMS in 2007 is a nationally representative survey (on 17,735 individuals divided in 5,575 households) conducted by the Statistical Office of the Republic of Serbia in cooperation with the World Bank. We have opted to use the LSMS dataset because it was the only comprehensive survey in Serbia providing the amount of social and economic information required for tax-and-benefit microsimulation modelling. Although we use the data from 2007, labour market indicators are similar to 2015 not only in absolute but also in relative terms, across both gender and levels of educational attainment. According to the Labour Force Survey Data for Serbia in 2015, unemployment and inactivity rates for women were 4 and 15 percentage points higher than for men, respectively, which is comparable to the situation in 2007. Further, those having primary education or less still have significantly higher inactivity rates compared 7 The main advantage of using discrete-choice instead of continuous labour supply models comes from the possibility of accounting for taxes and benefits (i.e., non-linear and non-convex budget sets; Van Soest, 1995), which is why these models have been extensively used for an ex-ante evaluation of hypothetical tax-and-benefit reforms. Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia 8 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 to individuals with secondary or tertiary education. Since the main labour market indicators (participation, employment, unemployment) were almost unchanged in 2015 compared to 2007, the dataset may still be regarded as relevant. The LSMS dataset contains separate data on formal and informal income from employment, self-employment and agriculture. According to the macro-validation results, formal employment incomes are lower than the actual amounts by 7%, with the difference being attributed to informal salaries and wages. In the case of agricultural and self-employment earners, the share of informal income in the actual income is much larger – 56% and 22%, respectively. Taxes, social contributions and benefits in SRMOD are simulated using the data on formal income, since both taxes and benefits are determined based on the formal income. Personal income tax, social security contributions and major means-tested benefits (monetary social assistance and child allowance) are the main tax-benefit programs simulated in SRMOD, while for most of the non-means-tested benefits (e.g., birth grant, old age pensions, unemployment benefits, caregiver allowance, maternity and childcare benefits) reported values are used, which is a common approach in other EUROMOD and related models, determined by the data availability.8 Personal income tax is simulated using the rules applicable in a year of the given dataset, which means that incomes from various sources are taxed at different, but always flat, tax rates ranging between 10% and 20%. Wages, as the largest source of income in Serbia, are taxed at the flat rate of 12%, applied to the amount of gross wages exceeding non-taxable threshold, which in 2007 amounted to RSD 5,050 per month.. Social security contributions are calculated on all gross labour incomes, applying the rate of 22% for pension and disability insurance contributions, 12.3% for health insurance and 1.5% for unemployment insurance and the regulations on the minimum base (35% of the average wage) and the maximum base (five times average wage) for social contributions. All social contributions are equally split among employers and employees. Monetary social assistance is the last-resort financial assistance program, means-tested against the total income (per household member), as well the land and buildings area (per household member) owned by the household. Thus, individuals or families who pass the means test are entitled to the benefit calculated using the following scale: i) for the first adult person in a family, the basic amount (RSD 7,628 per month) multiplied by 1; ii) for each additional adult person in a family, the basic amount multiplied by 0.5; and iii) for each child up to the age of 18, the basic amount multiplied by 0.3. Individuals incapable of work, and families with all members incapable 8 Detailed descriptions of the policies simulated in SRMOD are provided in Ranđelović and Žarković-Rakić (2013). of work, as well as lone-parent families with one or two children (below the age of 18) are entitled to an increased amount of this benefit (by 20%). Child allowance is the means-tested benefit aimed at reducing poverty in families with children. Eligibility is limited to the households in which total monthly net income per family member does not exceed a certain threshold (RSD 4,705 per month), while the amount of the benefit is flat (RSD 1,490 per month), with only the first four children in the family being entitled to this benefit. The results of macro-validation of SRMOD simulations are satisfactory, since the margin between the simulated income tax and social contributions compared to the administrative data ranges between 5% and 20%, a large share of discrepancy being attributed to the underreporting of income and sampling issues in the survey, since the average wage in the dataset is 8% lower than the one published by the Statistical Office (Ranđelović & Žarković-Rakić, 2013). The discrepancy between the simulated benefits and the amounts disclosed in administrative datasets is even lower, ranging from 5% to 15%. Using the data on market income, simulated taxes, social contributions and means-tested benefits, as well the reported (non-simulated) benefits, SRMOD provides calculations of the household disposable income, replacement rates, and effective marginal tax rates. In this paper, we estimate two discrete choice labour supply models, thus specifying separately preferences of singles and couples. Labour supply model estimation is restricted to the ‘labour market flexible’ individuals whose labour supply behaviour can be captured by the econometric model. Therefore, disabled individuals, students, pensioners, and self-em- ployed individuals are dropped, which is a common approach in the labour supply literature (Blundell et al., 2000; Figari, 2015; Haan & Myck, 2007). An additional reason to exclude the self-employed from the model is related to difficulties with measurement of their true hours and wages (Löfler et al., 2014). Descriptive statistics of the estimation sample are provided in the Appendix (Table A1). Since the model mostly deals with wage employment, similar to the literature on Southern European countries, it is focused on formal work, with the similar or slightly higher wage-employment informality (De Luca et al., 2013, Figari, 2015). Discrete choice labour supply models are based on the assumption that a household can choose among a finite number of J+1 working hours. Each hour j=0,...,J corresponds to a given level of disposable income of individual , and each discrete bundle of working hours and income provides a different level of utility. In other words, the utility of a household i making the choice j, Vij, is given by: Vij = U (Hfij,Hmij,Iij,Zi) + εij 15 distribution; this is important, since non-participation is extremely high among the low-income population. This means that IWB policies in Serbia would not only help reduce non-participation but also, to a certain degree, help reduce inequality, with net positive fiscal effects. The difference in the size of effects of IIWB and FIWB, depending on the income level and marital status, is not only the consequence of the difference in design of IIWB and FIWB policies but also the result of variation in labour supply elasticities by income levels and marital status. The results of this paper would certainly be important for informing the policymaking process in Serbia, as the government has recently started experimenting with policies similar to the IWB schemes. Given that the labour market structure and the design of the tax-and-benefit system in Serbia are quite similar to those of neighbouring countries, we believe that the results of our analysis could also be of interest to a wider range of economies in the region. The limitations of this paper, which also provide scope for further research, are as follows. First, this paper investigates the labour supply of IWB policy, while the labour demand analysis is beyond its scope. However, it is important to keep in mind that when there is involuntary unemployment, not all individuals who want to work are successful in finding a job. The employment effect of the IWB depends not only on the motivation of individuals to look for a job but also on the labour market capacity to accommodate them. In other words, the employment effects of IWB schemes would be also dependent on the state of the labour demand, suggesting that during times of economic prosperity there will be bigger employment gains after the introduction of the IWB than during crisis. There are studies that implicitly encompass the labour demand effects by using involuntary unemployment to describe the labour demand reaction (Bargain et al. 2010). However, use of this approach is limited to the datasets that provide the information on involuntary unemployment, which is not the case with the LSMS for Serbia in 2007. Second, the paper does not take into account general equilibrium effects. The paper by Kolm and Tonin (2011) does take such effects into account in a theoretical framework and shows how IWBs policies can be extended to larger sections of the workforce. Third, this type of analysis is inherently static; microsimulation was only used in an accounting manner (day after), and disposable income of a representative sample of the population was calculated before and after a reform using a tax-and-benefit calculator. Li and O’Donoghue (2013) provide an overview of the literature and the data requirements for the dynamic microsimulation models. Finally, although we justify in the paper why the policies simulated in SRMOD refer to the LSMS for Serbia in 2007 as the baseline year, the more recent SILC data for Serbia could be utilized, as in Žarković-Rakić et al. (2016) and Clavert et al. (2019). Acknowledgments This research has been generously funded by the Regional Research Promotion Program (RRPP) for the Western Balkans, supported by the University of Fribourg and the Swiss Development Agency, within the “Making work pay in the Western Balkans” project, conducted in cooperation between the Foundation for Advancement of Economics, Belgrade, Serbia and the University American College Skopje, Skopje, Macedonia. References Aaberge, R., Colombino, U., & Strøm, S. (1999). Labour supply in Italy: An empirical analysis of joint household decisions, with taxes and quantity constraints. Journal of Applied Econometrics, 14, 403-422. https://doi.org/10.1002/(SICI)1099-1255(199907/08)14:4<403::AIDJAE520>3.0.CO;2-4 Arandarenko, M., & Vukojević, V. (2008). Labor costs and labor taxes in the Western Balkans. In C. Bredenkamp, M. Gragnolati & V. Ramljak (Eds.), Enhancing efficiency and equity: Challenges and reform opportunities facing health and pension systems in the Western Balkans (pp. 119-160). Washington, DC: The World Bank. Arandarenko, M., Vladisavljević, M., & Žarković-Rakić, J. (2012). From inactivity to work: analysis of the public policies impact and factors affecting inactivity. In Cerović, et al. ed.: From Global Crisis to Economic Growth. Which Way to Take? VOLUME I, ECONOMICS. Faculty of Economics, University of Belgrade. Baldini, M., Bosi, P., & Toso, S. (2002). Targeting welfare in Italy: Old problems and perspectives on reform. Fiscal Studies, 23(1), 51–75. https://doi.org/10.1111/j.1475-5890.2002.tb00054.x Bargain, O., Caliendo, M., Haan, P., & Orsini, K. (2015). ‘Making work pay’ in a rationed labour market. Journal of Population Economics, 23, 323-351. Bargain, O., Orsini, K., & Peichl, A. (2013). Comparing labour supply elasticities in Europe and the United States: New results. Journal of Human Resources, 49(3), 723-838. https://doi.org/10.3368/jhr.49.3.723 Saget, C. (1999). The determinants of female labour supply in Hungary. Economics of Transition, 7(3), 575–591. https://doi. org/10.1111/1468-0351.00026 Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia 16 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 Bell, K. (2005). Tackling poverty and making work pay: Can tax credits do better? Presentation given at Institute for Fiscal Studies, 15 September 2005. Retrieved from http://www.ifs.org.uk/docs/bell_taxcredits05.ppt#5 Blundell, R. W., & Macurdy, T. (1999). Labour supply: A review of alternative approaches. In O. Aschenfleter & D. Card (Eds). Handbook of labour economics, edition 1, volume 3, chapter 27, pages 1559-1695. Amsterdam: Elsevier Science. https://doi.org/10.1016/S1573- 4463(99)03008-4 Blundell, R. W., Duncan A., McCrae J., & Meghir, C. (2000). The labour market impact of the working families tax credit. Fiscal Studies, 21(1), 75-103. https://doi.org/10.1111/j.1475-5890.2000.tb00581.x Blundell, R., & Hoynes, H. W. (2004). Has ‘in-work’ benefit reform helped the labor market? In Seeking a premier economy: The economic effects of British economic reforms, 1980-2000 (pp. 411-460). The University of Chicago Press, Chicago. https://doi.org/10.7208/ chicago/9780226092904.003.0011 Blundell, R. (2006). Earned Income Tax Credit policies: Impact and optimality. Labour Economics, 13, 423–443. https://doi.org/10.1016/j. labeco.2006.04.001 Brewer, M. (2006). Tax credits:fixed or beyond repair? In: Chote, R. et al. (Eds.), The IFSGreen Budget. London: Institute for Fiscal Studies. Brewer, M., Duncan, A., Shephard, A., & Suarez, M. J. (2006). Did working families’ tax credit work? The impact of in-work support on labour supply in Great Britain. Labour Economics, 13, 699-720. https://doi.org/10.1016/j.labeco.2005.11.002 Clavet, N.-J.,Tiberti, L.,Vladisavljević, M., Žarković-Rakić, J., Anić, A., Krstić, G., & Ranđelović, S. (2019). Reduction of child poverty in Serbia: Benefit or employment strategy? Economics of Transition and Institutional Change, 27,615–645.https://doi.org/10.1111/ ecot.12197 Colonna, F., & Marcassa, S. (2013). Taxation and labour force participation: The case of Italy. Bank of Italy Occasional Paper No. 191. https:// doi.org/10.2139/ssrn.2303708 Davies, P. (2018) Which countries have the worst income inequality in Europe?. Euronews. Retrieved from: https://www.euronews. com/2018/04/26/which-countries-have-the-worst-income-inequality-in-europe-. De Luca, G., Rossetti, C., & Vuri, D. (2013). In-work benefits for married couples: An ex-ante evaluation of EITC and WTC policies in Italy. IZA Journal of Labour Policy, 3:23. https://doi.org/10.1186/2193-9004-3-23 European Commission (2010). Europe 2020: A European strategy for smart, sustainable and inclusive growth. Retrieved from: https:// ec.europa.eu/eu2020/pdf/COMPLET%20EN%20BARROSO%20%20%20007%20-%20Europe%202020%20-%20EN%20version.pdf Eissa, N., & Hoynes, H. (1998). The Earned Income Tax Credit and the labour supply of married couples. NBER Working paper No. 6856. https://doi.org/10.3386/w6856 Eissa, N., & Liebman J. B. (1996). Labour tax supply response to the Earned Income Tax Credit. The Quarterly Journal of Economics, 111, 605-637. https://doi.org/10.2307/2946689 Eurostat (2019a). Employment Statistics. Statistics Explained. Retrieved from: https://ec.europa.eu/eurostat/statistics-explained/index. php/Employment_statistics Eurostat (2019b). Gini coefficient of equivalised disposable income-EU-SILC survey. Retrieved from: http://appsso.eurostat.ec.europa.eu/ nui/show.do?dataset=ilc_di12 . Figari, F. (2010). Can in-work benefits improve social inclusion in the Southern European countries? Journal of European Social Policy, 20(4), 301-315. Figari, F. (2015). From housewives to independent earners: Can the tax system help Italian women to work? Journal of Social Policy, 44(1), 63-82. https://doi.org/10.1017/S0047279414000555 Haan, P., & Myck, M. (2007). Apply with caution: Introducing UK-style in-work support in Germany. Fiscal Studies, 28, 43-72. https://doi. org/10.1111/j.1475-5890.2007.00047.x Hazans, M. (2011). Informal workers across Europe: Evidence from 30 countries. IZA Discussion Paper 5871. https://doi.org/10.1596/1813- 9450-5912 Heckman, J. (1976). The common structure of statistical models of truncation, sample selection and limited dependent variables and a simple estimator for such models. Annals of Economic and Social Measurements, 5, 475-492. Heckman, J. (1979). Sample specification bias as a selection error. Econometrica, 47, 153-162. https://doi.org/10.2307/1912352 Immervoll, H., & O’Donoghue, C. (2001). Welfare benefits and work incentives: An analysis of the distribution of net replacement rates in Europe using EUROMOD, a multi-country microsimulation model. EUROMOD working paper EM4/01. Immervoll, H., & Pearson, M. (2009). A good time for making work pay? Taking stock of in-work benefits and related measures across the OECD. IZA Policy Paper No 3. Immervoll, H. (2012). Activation policies in OECD countries – An overview of current approaches. OECD Social Protection and Policy Note. Koettl, J. (2010). Does formal work pay in Serbia? The role of labour taxes and social benefit design in providing disincentives for formal work. A technical note prepared for the World Bank. Working Paper. Washington, DC: World Bank. Kolm, A. S., & Tonin, M. (2011). In-work benefits and unemployment. International Tax and Public Finance, 18(1), 74-92. https://doi. org/10.1007/s10797-010-9148-6 Krstić, G., & Sanfey, P. (2011). Earnings inequality and informal economy: Evidence from Serbia. Economics of Transition, 19(1), 179-199. https://doi.org/10.1111/j.1468-0351.2010.00400.x Krstić, G., Arandarenko, M., Arsić, M., Radulović, B., Ranđelović, S., & Janković, I. (2015). Formalising the shadow economy in Serbia: Main findings and recommendations. In G. Krstić & F. Schneider (Eds.), Formalising the shadow economy in Serbia (pp. 123-158). Springer Open, New York. https://doi.org/10.1007/978-3-319-13437-6_9 17 Leibfritz, W. (2008). Increasing employment in FYR Macedonia: The role of labour taxation. A Policy Note prepared for the World Bank. Leppik, L. (2006). In-work benefits: Literature review. Tallin: PRAXIS Center for Policy Studies. Li, J., & O’Donoghue, C. (2013). A survey of dynamic microsimulation models: Uses, model structure and methodology. International Journal of Microsimulation, 6(2), 3-55. Löfller, M., Peichl, A., & Siegloch, S. (2014). Structural labour supply models and wage exogeneity. IZA Discussion Paper No. 8281. https:// doi.org/10.2139/ssrn.2464472 McFadden, D. (1974). Conditional logit analysis of qualitative choice behaviour. In P. Zarembka (Ed.), Frontiers in econometrics (pp. 105-142). New York, NY: Academic Press. Meghir, C., & Phillips, D. (2010). Labour supply and taxes. In J. Mirrlees (Ed.), The Mirrlees review. (pp. 202-274). Oxford: Oxford University Press. Meyer, B., &Rosenbaum, D. (2000). Making single mothers work: Recent tax and welfare policy and its effects. National Tax Journal, 53(4), 1027-1062. https://doi.org/10.17310/ntj.2000.4S1.02 Mojsoska-Blazevski, N. (2012). Taxation of labour: The effect of labour taxes and costs on employment in Macedonia. Post-Communist Economies, 24(2), 241-256. https://doi.org/10.1080/14631377.2012.675158 Mojsoska-Blazevski N., Petreski, M., &Petreska, D. (2015). Increasing the labour market activity of the poor, females, and informal workers: Let’s make work pay in Macedonia. Eastern European Economics, 53(6), 466-490. https://doi.org/10.1080/00128775.2015. 1103656 Myck, M., Kurowska, A., & Kundera, M. (2013). Financial support for families with children and its trade-offs: Balancing redistribution and parental work incentives. Baltic Journal of Economics, 13(2), 61-85. https://doi.org/10.1080/1406099X.2013.10840533 OECD (2009). A good time for making work pay? Taking stock of in-work benefits and related measures across the OECD. OECD Working Papers No. 81. OECD (2019). Labour force participation rate. Retrieved from: https://data.oecd.org/emp/labour-force-participation-rate.htm. Orsini, K. (2006). Is Belgium making work pay? KU Leuven Working Paper. https://doi.org/10.2139/ssrn.934317 Orsini, K., & Bargain, O. (2006). In-work policies in Europe: Killing two birds with one stone? Labour Economics, 13(6), 667-693. https:// doi.org/10.1016/j.labeco.2005.10.007 Owens, J. (2006). Fundamental tax reform: An international perspective. National Tax Journal, 59(1), 131–64. https://doi.org/10.17310/ ntj.2006.1.06 Ranđelović, S., & Žarković-Rakić, J. (2013). Improving working incentives: Evaluation of tax policy reform using SRMOD. International Journal of Microsimulation, 6(1), 157-176. St Martin, A., & Whiteford, P. (2003). More jobs and better pay. OECD Observer No. 239, September 2003. Retrieved from http://oecdobserver.org/news/archivestory.php/aid/1077/More_jobs_and_better_pay.html Saez, E. (2002). Optimal income transfer programs: Intensive versus extensive labour supply responses. The Quarterly Journal of Economics, 117. 1039–73. https://doi.org/10.1162/003355302760193959 Scholtz, J. K. (1994). The Earned Income Tax Credit: Participation, compliance, and anti-poverty effectiveness. National Tax Journal, 47, 59–81. Scholtz, J. K. (1996). In-work benefits in the United States: The Earned Income Tax Credit. Economic Journal, 106, 156–69. https://doi. org/10.2307/2234939 SORS (Statistical Office of the Republic of Serbia) (2016). Labour Force Survey in the Republic of Serbia, 2015, Bulletin 608, Belgrade: SORS. Sutherland, H., & Figari, F. (2013). EUROMOD: the European Union tax-benefit microsimulation model. International Journal of Microsimulation, 6(1), 4-26. Van Soest, A. (1995). Structural models of family labour supply: A discrete choice approach. Journal of Human Resources, 30, 63-88. https://doi.org/10.2307/146191 Žarković-Rakić, J., Ranđelović, S., & Vladisavljević, M. (2016). Labour market effects of Social Security contributions reform in Serbia. Economic Annals, 208, 73–92. https://doi.org/10.2298/EKA1608073Z Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia 18 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 APPENDIX Table A1. Descriptive statistics for the labour supply estimation mean std. dev. min max Secondary education 0.604 0.489 0 1 Tertiary education 0.154 0.361 0 1 Working experience 13.171 11.867 0 48 Settlement (Urban==1) 0.575 0.494 0 1 Region Vojvodina 0.257 0.437 0 1 Region West Serbia 0.306 0.461 0 1 Region East Serbia 0.262 0.440 0 1 Children under 3 years 0.125 0.373 0 3 Single 0.330 0.470 0 1 Age 40.359 11.917 18 64 Non-work hh income per adult equivalent (in 1.000 RSD) 3.332 5.537 0 75.80175 Total sample 6,473 Table A2. Wage equation for females and males, with Heckman correction Females Males  Coef. Std. Err. Coef. Std. Err. Hourly wage rate (ln)  Primary education (omitted) - - Secondary education 0.348*** (0.040) 0.207*** (0.035) Tertiary education 0.894*** (0.049) 0.698*** (0.045) Working experience 0.009*** (0.002) 0.003*** (0.001) Settlement (Urban==1) 0.185*** (0.027) 0.178*** (0.025) Region Belgrade (omitted) - - Region Vojvodina -0.192*** (0.037) -0.195*** (0.039) Region West Serbia -0.291*** (0.036) -0.226*** (0.036) Region East Serbia -0.348*** (0.038) -0.312*** (0.039) Constant 4.341*** (0.071) 4.708*** (0.058) Employment (1 = in employment) Primary education (omitted) - - Secondary education 0.323*** (0.070) 0.188*** (0.073) Tertiary education 0.909*** (0.089) 0.812*** (0.108) Working experience 0.096*** (0.006) 0.097*** (0.007) Settlement (Urban==1) 0.021 (0.054) -0.246*** (0.058) Region Belgrade (omitted) - - Region Vojvodina -0.145* (0.081) -0.030 (0.091) Region West Serbia -0.081 (0.077) -0.040 (0.086) Region East Serbia -0.120 (0.079) -0.035 (0.090) Children under 3yoa -0.528*** (0.078) 0.043 (0.076) Single 0.251*** (0.063) -0.130* (0.071) Age 0.173*** (0.021) 0.202*** (0.017) 19 Females Males  Coef. Std. Err. Coef. Std. Err. Age squared -0.003*** (0.000) -0.004*** (0.000) Non-work hh income per adult eq. (in 1.000 RSD) -0.016*** (0.005) -0.028*** (0.005) Constant -2.980*** (0.380) -2.661*** (0.344) Rho 0.33 (0.087) -0.0049 (0.090) Lambda 0.19 (0.046) -0.0025 (0.050) Sigma 0.57 (0.021) 0.51 (0.023) Observations 3,430 3,043 Censored N 1733 802 Wald test: joint significance [Chi2 (5)] 549.61 506.93 Prob > Chi2 0.000 0.000 LR test of indep. eqns. (rho = 0): 15.7 0.0030 Prob > chi2: 0.000 0.96  Table A3a. Preference estimates for singles (Conditional Logit) Total Females Males Coef. Std. Err. Coef. Std. Err. Coef. Std. Err. Income 5.186*** (0.715) 5.145*** (0.966) 5.477*** (1.174) *Age -0.221*** (0.038) -0.211*** (0.055) -0.248*** (0.061) *Age square (/100) 0.270*** (0.047) 0.271*** (0.073) 0.298*** (0.074) *Secondary ed.(a) -0.522*** (0.156) -0.802*** (0.259) -0.254 (0.247) *Tertiary ed. -0.461*** (0.130) -0.751** (0.382) -0.035 (0.280) * Children .(b) 0.112 (0.199) -0.049 (0.241) 0.431 (0.467) Square -0.017*** (0.002) -0.021*** (0.004) -0.024*** (0.009) Income* Hours of work 0.002* (0.001) 0.002 (0.001) 0.003 (0.002) Hours of work -0.532*** (0.029) -0.540*** (0.039) -0.529*** (0.048) *Age 0.017*** (0.002) 0.016*** (0.002) 0.019*** (0.003) *Age square (/100) -0.022*** (0.002) -0.021*** (0.003) -0.024*** (0.003) *Secondary ed. .(a) 0.027*** (0.006) 0.037*** (0.010) 0.015 (0.010) *Tertiary ed. 0.021*** (0.007) 0.021 (0.021) 0.008 (0.017) * Children.(b) -0.003 (0.022) 0.017 (0.030) -0.048 (0.039) Square 0.005*** (0.000) 0.005*** (0.000) 0.004*** (0.000) Fixed costs - - - * Children -0.578 (0.758) -0.586 (1.115) -0.187 (1.130) N (c) 1,992 3,231 2,745 Pseudo R Square 0.303 0.312 0.312 Log-likelihood -1525 -814.6 -691.1 Wald test: joint sig [Chi2 (16)] 1327.35 628.13 737.24 Prob > Chi2 0.000 0.000 0.000 Notes: (a)Primary education omitted .(b)Dummy variable for single family with child .(c)Estimated on the total of 5,976 observations = number of singles in the sample (1,992) multiplied by number of choices in simulation (3) Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia Table A2. Wage equation for females and males, with Heckman correction (continued) 20 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 Table A3b. Preference estimates for couples (Conditional logit) Total Female Male Coef. Std. Err. Coef. Std. Err. Coef. Std. Err. Income -0.358 (0.665) *Age 0.045 (0.034) -0.040 (0.035) *Age square (/100) -0.041 (0.040) 0.059 (0.040) *Secondary ed.(a) -0.120* (0.069) -0.120* (0.069) *Tertiary ed. -0.057 (0.095) -0.057 (0.095) * Children .(b) 0.452*** (0.112) Square -0.004*** (0.001) Income * Hours of work 0.001 (0.000) 0.000 (0.001) Hours of work -0.388*** (0.038) -0.286*** (0.040) *Age 0.009*** (0.002) 0.009*** (0.002) *Age square (/100) -0.013*** (0.002) -0.013*** (0.002) *Secondary ed. .(a) 0.028*** (0.005) -0.007 (0.005) *Tertiary ed. 0.039*** (0.009) -0.021** (0.009) * Children.(b) -0.037*** (0.006) -0.021*** (0.006) * Female and male hours Interaction (/100) 0.033*** (0.010) Square 0.005*** (0.000) 0.004*** (0.000) Fixed costs - - * Children -0.043 (0.247) 0.194 (0.213) N (c) 1,543 Pseudo R Square 0.346 Log-likelihood -2218 Wald test: joint sig. [Chi2 (30)] 2343.89 Prob > Chi2 0.000 Notes: (a)Primary education omitted .(b)Dummy variable for single family with child .(c)Estimated on the total of 13,887 observations = number of couples in the sample (1,543) multiplied by number of choices (9) Table A4a. Effects of the reforms for singles - percentages and number of people for each choice, total, first and fifth quintile Share of the choices Change (in pp) Number of people per choice Change no IWB with iIWB with fIWB iIWB fIWB no IWB iIWB fIWB iIWB fIWB Total Non-participation 42.1% 35.8% 33.3% -6.3** -8.8** 377,842 321,568 298,738 -56,274 -79,104 Part-time 7.3% 7.7% 8.7% 0.4 1.4 65,666 69,464 78,144 3,798 12,478 Full-time 50.6% 56.4% 58.0% 5.9** 7.4** 453,553 506,029 520,179 52,476 66,626 Female Non-participation 43.6% 37.5% 35.2% -6.0** -8.3** 187,292 161,362 151,411 -25,930 -35,881 Part-time 8.3% 8.5% 9.3% 0.2 1.0 35,632 36,565 39,987 933 4,355 Full-time 48.2% 54.0% 55.5% 5.8* 7.3** 207,116 232,113 238,641 24,997 31,526 Male Non-participation 40.8% 34.3% 31.5% -6.5** -9.3** 190,550 160,206 147,327 -30,344 -43,223 Part-time 6.4% 7.0% 8.2% 0.6 1.7 30,034 32,899 38,157 2,865 8,123 Full-time 52.8% 58.7% 60.3% 5.9** 7.5** 246,438 273,917 281,538 27,479 35,100 21 Share of the choices Change (in pp) Number of people per choice Change no IWB with iIWB with fIWB iIWB fIWB no IWB iIWB fIWB iIWB fIWB The 1st quintile Total Non-participation 99.3% 79.0% 73.3% -20.2** -26.0** 178,094 141,809 131,487 -36,286 -46,607 Part-time 0.6% 2.6% 4.4% 2.0 3.9 1,019 4,618 7,950 3,599 6,931 Full-time 0.2% 18.4% 22.3% 18.2** 22.1** 319 33,006 39,995 32,686 39,676 Female Non-participation 98.8% 79.2% 74.3% -19.6** -24.5** 83,379 66,842 62,673 -16,538 -20,706 Part-time 1.2% 3.1% 4.4% 1.9 3.2 1,019 2,625 3,680 1,605 2,661 Full-time 0.0% 17.7% 21.4% 17.7** 21.4** 0 14,932 18,045 14,932 18,045 Male Non-participation 99.7% 78.9% 72.4% -20.8** -27.3** 94,715 74,967 68,814 -19,748 -25,901 Part-time 0.0% 2.1% 4.5% 2.1 4.5 0 1,994 4,270 1,994 4,270 Full-time 0.3% 19.0% 23.1% 18.7** 22.8** 319 18,073 21,950 17,754 21,631 The 5th quintile Total Non-participation 14.7% 16.2% 14.6% 1.5 -0.1 26,254 28,882 26,048 2,628 -206 Part-time 9.0% 8.8% 9.4% -0.2 0.4 16,026 15,736 16,812 -290 786 Full-time 76.4% 75.0% 76.0% -1.3 -0.3 136,522 134,184 135,942 -2,338 -580 Female Non-participation 16.3% 18.6% 16.5% 2.4 0.2 15,380 17,626 15,597 2,246 217 Part-time 11.2% 11.0% 11.8% -0.3 0.5 10,635 10,379 11,153 -256 518 Full-time 72.5% 70.4% 71.7% -2.1 -0.8 68,585 66,595 67,850 -1,989 -735 Male Non-participation 12.9% 13.4% 12.4% 0.5 -0.5 10,874 11,256 10,451 382 -423 Part-time 6.4% 6.4% 6.7% 0.0 0.3 5,390 5,357 5,659 -33 268 Full-time 80.7% 80.3% 80.9% -0.4 0.2 67,937 67,589 68,092 -349 155 Notes: Stars in the table denote significant effects (* p<0.1; ** p<0.05); t-tests available upon the request. Data weighted by the weights provided by RSO. Table A4b. Effects of the reforms for couples - percentages and number of people for each choice, total, first and fifth quintiles Share of the choices Change (in pp) Number of people per choice Change no IWB with iIWB with fIWB iIWB fIWB no IWB iIWB fIWB iIWB fIWB Total Non-participation 39.1% 36.0% 38.3% -3.1* -0.8 525,866 484,883 515,241 -40,983 -10,625 Part-time 6.8% 7.3% 6.9% 0.4 0.1 91,751 97,687 92,754 5,935 1,003 Full-time 54.1% 56.7% -54.8% 2.7* 0.7 728,540 763,588 738,163 35,048 9,623 Female Non-participation 52.5% 49.5% 51.7% -3.0 -0.9 353,486 332,986 347,645 -20,500 -5,841 Part-time 6.4% 6.9% 6.5% 0.6 0.1 43,016 46,762 43,848 3,746 832 Full-time 41.1% 43.6% 41.8% 2.5 0.7 276,577 293,331 281,585 16,754 5,009 Male Non-participation 25.6% 22.6% 24.9% -3.0 -0.7 172,380 151,897 167,596 -20,483 -4,784 Part-time 7.2% 7.6% 7.3% 0.3 0.0 48,735 50,925 48,906 2,190 170 Full-time 67.1% 69.9% 67.8% 2.7 0.7 451,964 470,257 456,578 18,294 4,614 The 1st quintile Total Non-participation 83.9% 76.9% 82.1% -6.9** -1.8 228,222 209,333 223,433 -18,889 -4,789 Part-time 4.5% 5.0% 4.7% 0.4 0.1 12,348 13,487 12,724 1,139 376 Full-time 11.6% 18.1% 13.2% 6.5** 1.6 31,605 49,355 36,018 17,750 4,413 Female Non-participation 94.0% 88.4% 92.3% -5.6* -1.7 127,927 120,307 125,556 -7,620 -2,370 Part-time 1.2% 1.9% 1.5% 0.7 0.2 1,644 2,622 1,977 978 333 Full-time 4.8% 9.7% 6.3% 4.9 1.5 6,516 13,158 8,554 6,642 2,037 Saša Ranđelović, Jelena Žarković Rakić, Marko Vladisavljević, Sunčica Vujić: Labour Supply and Inequality Effects of In-Work Benefits: Evidence from Serbia Table A4a. Effects of the reforms for singles - percentages and number of people for each choice, total, first and fifth quintile (continued) 22 NAŠE GOSPODARSTVO / OUR ECONOMY Vol. 65 No. 3 / September 2019 Share of the choices Change (in pp) Number of people per choice Change no IWB with iIWB with fIWB iIWB fIWB no IWB iIWB fIWB iIWB fIWB Male Non-participation 73.7% 65.4% 71.9% -8.3* -1.8 100,295 89,026 97,876 -11,270 -2,419 Part-time 7.9% 8.0% 7.9% 0.1 0.0 10,704 10,865 10,747 161 43 Full-time 18.4% 26.6% 20.2% 8.2 1.7 25,088 36,197 27,464 11,108 2,376 The 5th quintile Total Non-participation 10.3% 9.5% 10.2% -0.7 -0.1 27,582 25,642 27,392 -1,940 -190 Part-time 8.4% 9.0% 8.5% 0.5 0.0 22,687 24,118 22,739 1,431 52 Full-time 81.3% 81.5% 81.4% 0.2 0.1 218,571 219,081 218,710 510 139 Female Non-participation 16.1% 15.0% 16.0% -1.1 -0.1 21,591 20,123 21,452 -1,469 -139 Part-time 10.5% 11.0% 10.5% 0.6 0.0 14,051 14,833 14,070 782 19 Full-time 73.5% 74.0% 73.6% 0.5 0.1 98,777 99,464 98,898 687 121 Male Non-participation 4.5% 4.1% 4.4% -0.4 0.0 5,991 5,519 5,940 -472 -51 Part-time 6.4% 6.9% 6.4% 0.5 0.0 8,636 9,285 8,669 649 33 Full-time 89.1% 89.0% 89.1% -0.1 0.0 119,794 119,616 119,812 -177 18 Notes: Stars in the table denote significant effects (* p<0.1; ** p<0.05); t-tests available upon the request. Data weighted by the weights provided by RSO. Ponudba dela in učinki ugodnosti zaposlenih na neenakost: ugotovitve za Srbijo Izvleček Nizka udeležba na trgu dela skupaj z visokoučinkovito davčno obremenitvijo nizkih ravni plač ustvarja plodna tla za vpeljavo ugodnosti zaposlenih v Srbiji. Naš članek ponuja vnaprejšnjo oceno dveh shem ugodnosti zaposlenih, ki sta usmerjeni k spodbujanju ponudbe delovne sile in bolj enaki porazdelitvi dohodka. Metodološki pristop kombinira mikrosimulacijski model na osnovi davkov in koristi z diskretnim modelom ponudbe dela. Naši rezultati kažejo, da bi lahko tako individualne kot družinske sheme ugodnosti zaposlenih znatno okrepile sodelovanje na trgu dela, četudi bi lahko družinske ugodnosti imele nespodbudne učinke na drugega prejemnika dohodka. Večina vedenjskih sprememb se zgodi med najrevnejšimi posamezniki z znatnimi redistributivnimi učinki. Ključne besede: ugodnosti iz naslova zaposlenosti, ponudba delovne sile, neenakost, model diskretne izbire, mikrosimulacija Table A4b. Effects of the reforms for couples - percentages and number of people for each choice, total, first and fifth quintiles (continued)