WPS5933 Policy Research Working Paper 5933 Severance Pay Compliance in Indonesia Vera Brusentsev David Newhouse Wayne Vroman The World Bank Human Development Networ Social Protection Unit January 2012 Policy Research Working Paper 5933 Abstract This paper contributes new evidence from two large payments. Widespread non-compliance is consistent with household surveys on the compliance of firms with trends in employment rigidity, which remained essentially severance pay regulations in Indonesia, and the extent to unchanged following the large increases in severance which changes in severance pay regulations could affect mandated by the 2003 law. These results suggest that employment rigidity. Compliance appears to be low, workers may benefit from a compromise that relaxes as only one-third of workers entitled to severance pay severance pay regulations while improving enforcement report receiving it, and on average workers only collect of severance pay statutes, and possibly establishing a 40 percent of the payment due to them. Eligible female system of unemployment benefits. and low-wage workers are least likely to report receiving This paper is a product of the Social Protection Unit, Human Development Network. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at dnewhouse@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Severance Pay Compliance In Indonesia Vera Brusentsev, David Newhouse, and Wayne Vroman1 JEL codes: J32, J65 Keywords: Severance Pay, Non-compliance, Indonesia, Labor Regulations 1 Brusentsev is Professor of Economics at Swarthmore College. Newhouse is a Labor Economist with the Social Protection Unit of the World Bank, and Vroman is a Senior Economist at the Urban Institute. This is a background paper for the Indonesian Jobs Report. We gratefully acknowledge Vivi Alatas, who provided indispensible leadership for the Jobs Report and was instrumental in adding of questions on severance pay to the Indonesian surveys. Lina Marliani and Ari Perdana provided invaluable research assistance. Finally, we thank Gordon Betcherman, Shubham Chaudhuri, Edgar Janz, Chris Manning, Pierella Paci, and Bill Wallace for helpful comments and support. 2 1. Introduction The financial crisis that emerged in the United States in mid-2007 quickly transformed into a global credit crunch that sharply reduced global trade flows. Many countries were affected by the severe global recession that followed, and millions of individuals throughout the world lost jobs, leading to increased informality and working poverty. In the aftermath of the crisis, policymakers throughout the world are questioning whether existing labor market regulations successfully balanced the competing goals of maintaining flexibility and protecting workers. While discussions over the scope of labor market regulations are often heated, views are typically based either on the predictions of theoretical models or anecdotes. The lack of evidence is particularly acute for developing countries, where hard data are rare and mechanisms for enforcing regulations tend to be limited. Lack of enforcement is a particularly serious concern, given that significant portions of workers report earning wages that fall below the statutory minimum.2 No systematic evidence, however, exists regarding compliance with other labor regulations, making it difficult to make an informed assessment of the costs and benefits of regulatory reform. This paper contributes new evidence on the compliance of firms with severance pay regulations in Indonesia, and the extent to which changes in severance pay regulations could affect employment rigidity. This evidence sheds light on a vigorous policy debate that arose in Indonesia prior to the onset of the global financial crisis, following the passage of a 2003 labor law that increased employment protection and severance pay. The debate pitted the business community, which warned that increases in severance pay regulations would reduce employment, against organized labor, which argued that severance pay makes an important contribution to income security. The reaction to proposed changes to severance pay regulations in 2006 exacerbated this debate. Following persistent lobbying by the business community, as well as firm surveys that showed that high severance pay obligations were a serious concern, the easing of severance pay requirements became a major part of investment- 2 See, for example, World Bank (2011) finds that 40 percent of wage workers report earnings below the minimum wage in Indonesia in 2007. This is broadly consistent with reported estimates that non- compliance rates are as high as 33 percent in Costa Rica (Gindling and Terrell, 2007) and 45 percent in South Africa (Bhorat et al, 2010). 3 climate reforms proposed by the coordinating ministry of economic affairs. These proposed changes, however, were put on hold in April 2006, after large demonstrations were organized by labor unions (Manning and Roesad, 2007). Given the controversy surrounding severance pay in Indonesia, an analysis of survey data can impart important information on the performance of the current severance pay regulations. We exploit two household surveys that contain information on the receipt of severance pay following job separation: The Indonesian Family Life Survey (IFLS), which collected data on severance pay for the first time in 2007, and the National Labor Force Survey (Sakernas) which collected data on severance pay for the first time in 2008. These surveys provide a unique opportunity to conduct the first household-level analysis of severance pay in Indonesia, and to our knowledge, the developing world (Holzmann, et al, 2011).3 The results show that compliance is low – only a third of workers entitled to severance pay report receiving it, and on average workers only collect 40 percent of the payment due to them. Eligible female and low-wage workers are least likely to report receiving payments. Widespread non-compliance is consistent with trends in employment rigidity, which essentially remained constant following the large increases in severance mandated by the 2003 law. Weak enforcement in practice suggests that workers may benefit from a compromise that relaxes severance pay regulations while establishing a system of unemployment benefits. Enforcement can also be improved by establishing an administrative entity dedicated to monitoring compliance with severance pay regulations can improve the existing system. Other countries with severance pay programs, such as Barbados and Slovenia, have established arrangements to oversee severance pay statutes. The structure of the paper is as follows. In section two, we review the concerns of both the business community and employee representatives and present evidence on aggregate trends in labor market rigidity. Section three turns to the survey data, and demonstrates that not only are few workers eligibile for severance pay, but most eligible workers do not report receiving benefits. The descriptive analysis is followed in section four by suggestions for policy reform. 3 Vodopivec, et al (2009) finds that one third of all covered workers do not receive severance pay, in the high-income context of Slovenia, MacIsaac and Rama (2001) find that Peruvian workers that are more likely to be covered earn lower wage, but they are unable to observe receipt of severance pay directly. 4 2. Severance Pay and Employment Rigidity The first concern about severance pay in Indonesia is that the costs associated with the program are high and, as a consequence, distort firms’ hiring and firing decisions. Business representatives argue that these high costs discourage them from hiring new employees on a permanent basis, influence their lay-off decisions in response to contemporaneous economic conditions and lead to a deterioration in the investment climate. Severance pay compensation rates in Indonesia began to increase in the mid- 1990s. There were three changes to these regulations in the last decade of the last century and he first decade of this century: 1996, 2000, and 2003. Figure 1 shows the severance pay rates associated with these three legislative changes to the regulations. In 1996, the severance payment was estimated to be equivalent to a “hiring tax� of about 2.0 monthly wages per employee.4 The legislative changes in 2000 led to an increase in the “hiring tax� to an average of 3.4 months. With the changes introduced in 2003, the average severance payment is about 4.1 monthly wages per employee or 34 percent of an employee’s annual wage. The maximum severance pay is close to 30 months of wages in the case of a dismissal for an economic cause beyond the control of the employee. The current employment protection legislation, Manpower Law (No. 13/2003), regulates severance pay rates in Indonesia. In terms of severance regulations, Indonesia falls into the group of economies in East and South Asia with more generous statutes, in the same range as Bangladesh, Korea, Nepal, the Philippines, and the People’s Republic of China. As stated above, business representatives in Indonesian claim that severance pay costs are high and constrain the workforce decisions that firms make. One way to evaluate these claims is to examine the trends in various indicators of labor market rigidity. Using unemployment data and work experience data from the annual Sakernas labor force surveys from 1999 to 2008 for turnover in the labor market, 4 The hiring tax measures the discounted expected cost, at a time a worker is hired, of dismissing him/her in the future or if the worker quits. It shows the severance payment an employer expects to pay in the future for hiring a new worker. The calculation is based on the probability a new worker leaves in a particular year based on the reason for the separation. Different rates apply depending on the specific reason for the separation (quit, dismissal for minor violations or economic cause). See Alihsjahbana (2007) and Padjadjaran University (2004). 5 Figure 2 shows the evolution of key indicators in Indonesia. The unemployment rate shows a slight upward trend over the last decade. Except for 2006, the three turnover rates are contained within a trendless band over the same period. The percentage of all labor market participants who experienced a job separation in the previous year is shown in Figure 2 as the turnover rate, which is an indicator of the degree of employment rigidity. Except for 2006, the turnover rate consistently lies in the range between 6.0 to 8.0 percent. The largest proportion of all job separations was voluntary. Of the involuntary job separations, lay-off or business failure was cited by the majority of individuals as the main reason for the job separation.5 With one exception, involuntary separations consistently range between 2.0 and 3.0 percent of annual employment. Of all persons who experienced a job separation, the majority were wage and salary employees working in the formal sector. For nine of the ten years, from 1999 to 2008, they consistently represent between 50 and 75 percent of all job separations. Except for 2006, the range of employee separations falls between 3.7 and 5.6 percent of annual employment. Overall, Figure 2 indicates that, despite the legislative changes to severance pay regulations, turnover rates have been constant. Hence, the data do not support the claim that labor market flexibility has been compromised by the increases in severance pay introduced with the legislative changes. It is important to note that only a small share of the working population in Indonesia is covered by severance pay regulations because most labor market participants are employed in the informal sector. Of formal sector employees, most job separations tend to be voluntary. The majority of formal sector employees who do separate from their jobs are eligible for severance. Few of these employees, however, collect their full entitlement to severance pay. Actual practice departs substantially from the statutory regulations; the costs of severance pay are much lower than those implied by the severance statutes. As a result, 5 Similarly, voluntary separations form the majority of responses in the IFLS (2007) survey. Involuntary job separations accounted for 21.8 percent of all separations and these respondents cited two main reasons: company closed/relocated/restructured; fired for other reason. 6 the actual costs of severance pay based on data from the two surveys examined in this paper are 10 to 14 percent of the costs suggested by the statutory regulations. Three possibilities could account for the discrepancy: employees are not aware of the statutory regulations; compliance by firms is low; enforcement of the regulations relating to severance pay appears to be low in Indonesia. It seems likely that employers avoid paying severance by signing short-term contracts with employees, some of whom are unaware of their right to severance pay. Other employees may know of their rights but avoid involvement in the legal process to enforce their right due to delays and litigation costs. This paper does not examine these issues because we do not have the requisite data on knowledge, costs and delays. Gathering such information is an important direction for future research and could be achieved by adding supplementary questions into the surveys. While the data do not permit us to directly examine the enforcement of severance pay regulations in Indonesia, we can explore the first two possibilities. Both surveys can directly address the concern raised by employee representatives: compliance of severance pay regulations by firms is low. Prior to the introduction of a set of survey questions about severance pay, the claims of employee representatives were speculative in nature. The two surveys that collected data on severance pay for the first time in 2007 and 2008 give us the opportunity to test the validity of this claim. The following section presents information on the scope of severance pay and compliance of pay regulations. 3. Coverage and Scope of Severance Pay The Indonesian Family Life Survey (IFLS) collected data on severance pay for the first time in 2007, while the National Labor Force Survey (Sakernas) collected data for the first time in 2008. Three aspects of severance pay are described in the analysis below: coverage of and eligibility for severance pay; the scope of severance pay; and the compliance of current severance pay regulations. 3.1 Coverage of and eligibility for severance pay Data from the work experience modules of the two surveys are used to address the question of who is more likely to experience a job separation. As wage-and-salary labor 7 market participants are potentially covered by severance pay regulations, the analysis concentrates on those government and private sector employees who experienced a job separation. For the IFLS (2007) survey, the job separation refers to the previous five years of employment; to the previous two years in the Sakernas (2008) survey. The specific reason for the job separation is important because it affects both the entitlement to compensation and the size of the prospective payment. For instance, if a job separation is due to an economic cause beyond the control of the employee, the compensation package (both severance pay and long-service leave) is significantly larger and the cost to the firm is greater. In the IFLS (2007) survey, involuntary separations (lay-off and business closure) accounted for 21.8 percent of all job separations; voluntary separations were 56.3 percent and “other� accounted for 21.9 percent. The responses identifying the reason for a separation are similar in the Sakernas (2008) survey. Involuntary separations (lay-off and business failure) accounted for 27.9 percent of all job separations. But 39.4 percent of all responses were due to “other� reasons where the initiative for the separation cannot be identified. Not knowing the reason for the separation makes it difficult to determine the potential severance payment. As stated above, most of the job separations in both of the surveys are voluntary. Table 1 compares job separators with all government and private sector employees using the Sakernas (2008) survey. Employees who experienced a job separation in the previous two years are disproportionately younger, female, and well-educated. Job separations are concentrated in industries where entry and exit is relatively easy and mainly in the private domestic sector. Job separators have a relatively short employment relationship with the firm and tend to earn lower monthly wages. The turnover rate is higher for younger age groups and steadily decreases with age (except for the oldest group aged 55 years and older). While there are about twice as many men employed in the formal wage-and-salary sector, the turnover rate for women is about two-thirds higher. Job separators are relatively well-educated and the turnover rate is highest for those who graduated from high school. There is a disproportionate concentration of job separations in industries where entry and exit is relatively easy, such as manufacturing, trade, and especially public 8 services. Moreover, job separators have a relatively short employment relationship with the firm (that is, their tenure is less than four years). Knowing the length of tenure is important in assessing the potential costs associated with the severance regulations, particularly since Manpower Law 13/2003, Article 156, raised the rates of severance pay for employees with three or more years of tenure.6 The majority of job separators worked in the domestic private sector either for a domestic firm (36.4 percent) or as a domestic worker (45.8 percent).7 There was little difference, however, in the size of the firm in which they previously worked. Those persons who separated from their job tend to be employed in agriculture, trade and social services where the turnover rates are 21.5 percent, 11.9 percent, 21.2 percent, respectively. The firms in which all job separators were employed tend to be concentrated in Bali and Java.8 Most of the turnover occurs amongst employees with low job tenure. While the proportion of all wage and salary employees with tenure between 0 – 3 years is 0.4660, it is 0.7489 for employees with a job separation. The turnover rate for those with tenure between 0 – 3 years is 19.5 percent. Compared to all wage-and-salary employees, job separators tend to earn lower monthly wages. In addition, they tend to be concentrated at the lower end of the wage distribution and face a higher turnover rate. 3.2 Scope of severance pay Employees who were eligible for severance pay and received a payment after a job separation tend to be young, fairly well-educated and working in large firms in industrial regions of the country. The data indicates that few are aware of their entitlement to severance pay. This lack of awareness may be due to the complexity of the legislation regarding severance pay but, more than likely, it is exacerbated by the lack of formal contracts that explicitly define the employment relationship. 6 See figure 4.4 in World Bank (2011) for a detailed description of severance pay regulations 7 Question R11a of the Sakernas (2008) survey asks individuals to identify the type of business concern where they worked before the job separation occurred: government agency; state-owned firm; domestic private firm; foreign firm; and domestic worker. 8 The questions in the firing module of Sakernas (2008) pertain only to the industrial districts of Indonesia. 9 In the IFLS (2007) survey data, approximately 35 percent of all labor market participants work in the formal sector. These wage-and-salary employees are potentially covered by the severance regulations. Of those individuals in covered employment who are employed by a firm falling under the auspices of the regulations, not all meet the requirements. Furthermore, of those who do meet the necessary coverage requirements, not all are eligible to receive severance pay.9 Of government and private sector employees working in covered employment, only 14.8 percent state that they are entitled to severance pay as one of the set of benefits offered by their employer. Based on this low percentage, it seems that many formal sector employees lack a clear understanding of severance pay eligibility. In both surveys, women in covered employment are less likely than men to be eligible for severance pay in the event of a job separation.10 As Table 2 shows, eligibility for severance pay increases strongly with age, but there are no noticeable differences in eligibility by level of education. While there is variation in eligibility for severance pay across industries, there is no apparent pattern. There is, however, a strong systematic relationship between eligibility for severance pay and tenure: eligibility increases strongly with tenure. For firms of different sizes, the eligibility proportion steadily increases with firm size. There are two noticeable differences, however, between the surveys. The first difference is the likelihood of eligibility for severance across firm ownership status. In the IFLS (2007), wage and salary workers employed by state-owned enterprises are more likely to be eligible for severance than employees working in a government agency. On the other hand, the likelihood of eligibility for severance is similar across firm-ownership status in Sakernas (2008). Second, neither survey shows a consistently positive association between eligibility and the level of wages across the full range of wage rates. 3.3 Compliance with current severance pay regulations 9 Question TK25A3 of the IFLS (2007) survey asks individuals to identify a set of benefits offered by the employer. One of these benefits is severance payment eligibility. 10 Table 2 display details of severance eligibility and receipt of severance pay based on the Sakernas (2008) survey. The IFLS (2007) findings are consistent. 10 While payment of severance and long-service leave after a job separation is mandated by legislation in Indonesia, information from survey data suggests that there is a high rate of non-compliance with the regulations. Approximately one-third of legally eligible wage-and-salary employees receive a severance payment after a job separation occurs; depending on the survey, the average ratio of the severance payment received to the legally entitled amount is either slightly over one-third (IFLS, 2007) or just under one-third (Sakernas, 2008). There are two aspects of non-compliance. The first aspect is the proportion of eligible formal sector employees who receive a payment. The second aspect is the proportion of the entitlement amount that is actually received. The receipt of severance pay depends on two factors: the proportion of job separators who are eligible (as described in the previous section) and the proportion of those eligible who actually receive a payment after a job separation occurs. Both factors determine the ratio of actual severance payment to the legally entitled amount. The present discussion emphasizes this ratio. In both surveys, the proportion of eligible employees who received severance pay increases with age but is essentially the same for individuals with differing levels of formal education. In the IFLS (2007) survey, there is no difference between women and men in terms of severance pay receipt. In the Sakernas (2008) survey, however, eligible men are about 39 percent more likely to receive severance than eligible women. There is variation in the share of eligible persons who received severance pay by industry in both surveys, but there is no apparent pattern. There is a strong systematic relationship between the share of eligible employees who received severance pay and tenure: the share increases with the years in continuous employment with the same firm. Similarly, there is a strong systematic association between the share of eligible employees who received severance pay and the size of the firm: the proportion steadily increases with firm size. About one-quarter of eligible wage-and-salary workers receive severance pay if they are employed by a government agency. Eligible wage-and-salary workers employed by foreign or transnational corporations are the most likely to receive severance pay, while eligible workers employed by state-owned enterprises or domestic private firms are equally likely to receive severance pay. There is more variation in 11 receipt of severance pay among eligible employees across firm ownership, however, in the Sakernas (2008) survey. Generally, eligible employees who received severance pay steadily increases with the wage level in both surveys. While only a limited number of job separators actually received severance pay, those who did were generally satisfied with their payment. Only the IFLS (2007) survey specifically addresses employee satisfaction with respect to severance pay. Of the four levels of satisfaction identified in the survey, 68.6 percent stated that they were either very satisfied or satisfied with their compensation while only 28.9 percent indicated they were dissatisfied and just 2.5 percent stated that they were very dissatisfied with the severance pay they received. On the first aspect of non-compliance, the proportion of eligible formal sector employees who receive a payment, two-thirds of all eligible employees who separated from a job did not receive their severance payment. The second aspect is the proportion of the entitlement amount that is actually received. It is to this aspect that we now turn. The conclusion we draw from the survey data is that employers who make a severance payment actually pay less than the amount to which their employees are entitled. Table 3 shows the amount of legally entitled severance pay and actual payments by a set of employee and job characteristics. These calculations are based on weighted data from the IFLS (2007) survey. The mean monthly wage among recipients was 968,892 IDR; the severance entitlement amounted to 9,361,576 IDRs of which about 40 percent had actually been received to date (3,607,587 IRDs). Part of both totals, however, could be associated with long-service compensation as well as severance pay. The average amount actually received by women and men as a proportion of the legal entitlement is nearly identical: 0.3937 and 0.3816, respectively. Both legally entitled severance pay amounts and actual severance pay received increases with age. The proportion of actual receipt to the legally entitled severance pay ranges from 0.1138 at ages 15-18 years to 0.8883 at ages 50-64 years. These proportions compare to the all- age average of 0.3854. The legally entitled amount and the proportion of that amount received both increase with the level of education except for the highest group, those with tertiary education. 12 Generally, there is variation in both legally entitled and actual severance pay received across industries, but there is limited deviation from the overall average when the ratio of the two is considered. There are, however, important exceptions. Three industries demonstrate below average ratios of actual severance pay received to legally entitled payments: financial services, construction, and manufacturing. In terms of tenure, the relationship between both legally entitled and actual severance pay is highly positive and systematic. Legally entitled severance pay amounts increase from 0.923 million IDRs for formal sector workers with less than one year of tenure to 22.154 million IDRs for employees with 10+ years. The ratio of actual to legally entitled severance pay increases with tenure, rising dramatically for long-tenured employees. For all groups with tenure of less than 10 years, the ratio is less than half the overall average (0.3854) whereas for tenure of 10+ years it is 0.6169, 1.6 times the overall average. It is still low, however, for the 10+ year group. For firms of different sizes, there is no systematic relationship between both legal entitlement to and actual severance pay received. In firms with fewer than 100 employees, the ratio of actual severance pay received to the legally entitled amount is below the overall average of 0.3854; on the other hand, in the largest firms the ratio is 0.5814, 1.5 times the overall average. Wage-and-salary workers employed by government agencies are the least likely to receive the severance payment to which they are legally entitled, less than half the overall average. Employees in state-owned enterprises are the most likely to receive the full amount to which they are legally entitled. Employees in domestic private firms have approximately the average likelihood of recipiency while those in foreign or transnational corporations have a likelihood of nearly 50 percent above the overall average. Across the wage distribution, there is a systematic relationship between both legal entitlement to and actual severance pay received. While the legally entitled amount is approximately ten times the monthly wage for all-wage groups, those with below-average wages receive a below-average share of their legal entitlement. Except for the highest group receiving a monthly wage of more than 1.5 million, for all other employees the ratio is below the overall average. For the high wage group, however, the likelihood of receipt is 1.4 times the overall average. 13 Table 4 shows the amount of legally entitled severance pay and actual payments by a set of employee and job characteristics using data from the Sakernas (2008) survey. Here, the amount actually received by women as a proportion of their legal entitlement is higher than that for men: 0.3559 and 0.2697, respectively. Generally, both legally entitled severance pay amounts and actual severance pay received increase across the age distribution up to 35-49 years: from 0.0681 at ages 15-18 years to 0.3985 at ages 35-49 years. Contrary to a priori expectations, the proportions for the oldest two age groups drastically decline. These proportions compare to the all-age average of 0.2961. Except for those employees with less than primary education, the legally entitled amount increases with education. The ratio of the amount actually received, however, only increases at the highest education levels (that is, senior secondary and tertiary). Generally, there is wide variation in both legally entitled and actual severance pay received across industries. The industry patterns differ greatly across the two surveys. The relationship between both legally entitled and actual severance pay is positive and systematic for tenure. Legally entitled severance pay amounts increase from 1.135 million IDRs for those with less than one year of tenure to 22.942 million IDRs for employees with 10+ years. Generally, the ratio of actual to legally entitled severance pay also increases with tenure, rising from 0.1897 for less than one year to 0.3055 for employees with 10+ years of tenure. For firms of different sizes, there is a positive and systematic relationship between both legal entitlement to and actual severance pay received. In firms with fewer than 20 employees, the ratio of actual severance pay received to the legally entitled amount is below the overall average of 0.2916; on the other hand, in the largest two firm-size categories, both ratios exceed this average. Wage-and-salary workers employed by government agencies are the least likely to receive the severance payment to which they are legally entitled, only about 20 percent of the overall average. Employees in domestic private firms are the most likely to receive the full amount to which they are legally entitled: 0.3816. Employees in state-owned enterprises have approximately the average likelihood of recipiency while the likelihood in the remaining three firm-status categories is below-average. Across the wage 14 distribution, there is no apparent relationship between both legal entitlement to and actual severance pay received. If a job separation is due to an economic cause beyond the control of the employee, the compensation package (both severance pay and long-service leave) is significantly larger and the cost to the firm is greater. Hence, it is important to identify those employees who were fired for an economic cause. In the IFLS (2007) survey 16.3 percent of wage and salary employees with a job separation were fired in the past five years because the firm merged or restructured. While the proportion of these job separators who were legally eligible to receive severance pay on termination was 0.8807, only 0.4041 actually received a payment.11 The average severance entitlement amounted to 11,827,577 IDRs and the average amount of severance paid was 4,864,027 IDRs; about 40 percent of the amount to which an employee dismissed for an economic reason was entitled. This proportion of severance pay actually received (0.4112) was similar to the overall average of 0.3853 for all employees legally entitled to severance payments. Given that a job separation due to an economic cause beyond the control of the employee would automatically confer eligibility, the similarity in both proportions suggests a high degree of non-compliance. In the Sakernas (2008) survey, 16.0 percent of employees were separated from their job in the last two years because they were laid-off by their firm; approximately one-third were laid-off because the company failed.12 This proportion represents 4.2 percent of all job separations in the previous two years. While the proportion of these job separators who were legally eligible to receive severance pay on termination was 0.7859, only 0.1607 actually received it. The average severance entitlement amounted to 45,256,847 IDRs and the average amount of severance paid was 15,646,787 IDRs; about 34.6 percent of the amount to which an employee dismissed for an economic reason was entitled. This proportion of severance pay actually received (0.3457) was similar to the 11 Employees are legally eligible if (1) they have a work contract or letter of appointment which is permanent/open-ended; (2) dismissed for reasons stipulated by legislation. 12 Question R18 of the Sakernas (2008) survey asks wage and salary employees who were laid- off to identify the main reason for the lay-off. The proportion who responded that their job separation was due to the rationalization of the firm is 35.8 percent; followed by 26.0 percent for company bankruptcy/loss/disaster. 15 overall average of 0.2916 for all employees legally entitled to severance payments. Similar to the IFLS (2007), both proportions suggests a high degree of non-compliance. The results of the analysis highlight two important findings about the compliance of employers with the severance pay regulations. First, employees who are more vulnerable are more affected by non-compliance. Second, while de jure severance costs may be high for covered wage-and-salary workers, firms frequently do not actually pay the legally entitled amount to eligible employees. The required payments may not be delivered when the payment is due, however, if a firm enters bankruptcy because other creditors may have higher priority claims to the assets of the firm. If the rate of non- compliance due to bankruptcy is high, then an employer-based severance program would be ineffective in providing income security to wage-and-salary employees. A sound company, on the other hand, may face above-average costs when the volume of severance payments is high. 4. Conclusions The current impasse in reforming severance regulations reflects objections from both employers and employees. While recent negotiations have either collapsed or ended in a stalemate, renewed efforts are needed to find a workable compromise between the interests of employers and those of employee representatives. In addition, the reform options need to consider the voice of vulnerable workers who have been excluded from the debate. While it is not possible to assess the degree of hardship wage-and-salary workers face following a job separation, it is possible to infer that workers experience two significant disadvantages. First, severance payments are received by only a minority of eligible employees. Second, the severance payments to recipients are only a fraction of the legally-entitled amounts. Approximately one-third of legally-eligible employees actually receive severance pay after a job separation. On average, the ratio of the severance pay received to the legally-entitled amount is below 40 percent. The product of these two ratios yields a wage-loss protection share of between 10 to 14 percent of eligible severed workers’ monthly wages. Indonesia lacks an administrative structure to pool the risk of a job separation between employers and employees. While the statutory severance pay regulations appear 16 to be high, current employer practices contain the costs for firms, especially for low-wage earners. Our recommendation is to find a workable compromise that not only balances employer concerns about high severance pay rates and employee concerns about income security, but also considers vulnerable workers who have been excluded from the debate. It would be appropriate for Indonesia to establish an administrative entity to monitor compliance with the newly-negotiated severance pay regulations. Other countries with severance pay programs, such as Barbados and Slovenia, have established arrangements to oversee severance pay statutes. Indonesia may want to follow this lead. Regardless of the policy pursued by the Indonesian government, further data collection and research are needed to assess the impact of severance pay or evaluate future proposed programs. Currently, the data available on severance pay in national surveys are limited. Both the IFLS and Sakernas have collected data for only one year (2007 and 2008, respectively). Continuing collection of severance pay is vital for examining the labor market outcomes associated with changes in severance pay regulations, especially the employment and wage effects for workers in the formal sector and those in the informal economy. Additional research is needed to understand the barriers to compliance: lack of knowledge, lack of enforcement mechanisms, costs of legal recourse, and how firms avoid compliance. 17 Figure 1 m onths of s alary 30 25 20 15 10 5 0 <1 3 5 10 20 M ax Ye ars of s e rvice 1996 law 2000 law 2003 law Source: LP3E FE UNPAD and GIAT, 2004 18 Figure 2 Unemployment rate and turnover indicators from 1998 to 2007 14 12 10 Percent 8 6 4 2 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Unemployment rate Turnover Involuntary separations Employee separations 19 Table 1 Comparison of employee and job characteristics Percent of Percent of Turnover Percent of Turnover population wage a rate: separations rate: that are wage employees all that are involuntary employees with job separations involuntary separations separation in pat two in past two years years Age in years 15–19 6.4% 11.7% 21.8% 10.0% 4.7% 20–24 17.6% 30.6% 21.1% 25.5% 4.4% 25–34 34.0% 37.5% 13.0% 39.4% 3.4% 35–44 24.1% 12.4% 6.6% 15.7% 2.1% 45–54 13.7% 4.8% 4.5% 6.7% 1.6% 55 and older 4.2% 3.0% 9.3% 2.7% 2.1% Average 34.4 years 28.4 years 10.5% 29.7 years 2.8% Gender Female 33.3% 45.3% 15.3% 62.6% 5.3% Male 66.7% 54.7% 10.6% 37.4% 1.8% Education Less than primary 5.9% 4.7% 10.3% 5.0% 2.8% Primary 20.7% 17.4% 12.6% 20.3% 3.7% Junior secondary 42.7% 21.8% 6.6% 45.8% 3.5% Senior secondary 16.8% 42.5% 26.7% 22.5% 3.5% Tertiary 13.9% 13.6% 9.7% 6.5% 1.2% Industry Agriculture 9.1% 5.7% 21.5% 4.8% 4.6% Mining 1.6% 0.2% 4.3% 0.1% 0.4% Manufacturing 25.7% 15.6% 6.0% 20.3% 2.0% Utilities 0.6% 0.0% 0.8% 0.0% 0.0% Construction 6.2% 2.8% 6.9% 2.1% 1.3% Trade 14.8% 16.9% 11.9% 15.1% 2.7% 20 Transportation 6.4% 4.3% 8.6% 4.6% 2.3% Financial services 3.9% 3.5% 7.8% 1.9% 1.1% Public services 31.7% 51.0% 21.2% 51.0% 5.3% Region Sumatera 17.2% 8.3% 12.0% 5.3% 1.9% Java/Bali 69.7% 89.6% 12.4% 93.7% 3.2% Kalimantan 5.1% 0.8% 14.7% 0.6% 2.9% Sulawesi 4.9% 1.4% 10.1% 0.4% 0.7% Tenure 0–3 years 46.6% 74.9% 19.5% 63.4% 4.1% 4–9 years 25.9% 16.5% 7.8% 23.9% 2.8% 10+ years 27.5% 8.6% 4.0% 12.7% 1.5% Wage (IDRs) Below 250,000 7.0% 7.5% 19.1% 5.7% 3.6% 250,001–500,000 19.6% 22.1% 15.8% 17.6% 3.2% 500,001–1 mil. 36.5% 45.7% 14.6% 51.9% 4.2% 1–1.5 million 16.1% 14.3% 10.1% 16.4% 2.9% Above 1.5 mil. 20.8% 10.5% 5.8% 8.4% 1.2% Average 1,126,790 940,404 990,303 Source: Sakernas (August 2008) survey. 21 Table 2 Severance eligibility and receipt by employee and job characteristics Employees eligible Employees who Employees who for severance pay as received severance received severance a proportion of all pay as a proportion of pay as a proportion job separations all eligible of all separations Age 15–18 years 16.1% 12.9% 2.1% 19–24 years 16.2% 13.4% 2.2% 25–34 years 17.3% 39.5% 6.8% 35–49 years 26.1% 56.6% 14.8% 50–64 years 20.8% 36.6% 7.6% 65 years and older 29.7% 66.4% 19.7% All ages 18.4% 33.4% 6.1% Gender Female 14.9% 26.8% 4.0% Male 21.3% 37.2% 7.9% Education Less than primary 20.6% 25.1% 5.2% Primary 23.9% 30.7% 7.4% Junior secondary 19.1% 34.6% 6.6% Senior secondary 18.0% 36.5% 6.6% Tertiary 8.8% 28.2% 2.5% Industry Agriculture 19.2% 24.6% 4.7% Mining and quarrying 17.6% 0.0% 0.0% Manufacturing 24.4% 32.4% 7.9% Utilities 0.0% N/A N/A Construction 24.3% 58.3% 14.2% Trade 13.6% 43.8% 6.0% Transportation 21.2% 72.7% 15.4% 22 Financial services 15.0% 15.5% 2.3% Public services 17.8% 27.5% 4.9% Tenure Less than 1 year 14.1% 9.6% 1.4% 1–5 years 16.4% 25.9% 4.3% 5–10 years 26.8% 61.7% 16.5% 10+ years 34.2% 60.9% 20.8% Firm size 1–4 employees 10.1% 18.8% 1.9% 5–19 employees 14.0% 26.1% 3.7% 20–99 employees 21.3% 37.6% 8.0% 100+ employees 25.4% 37.5% 9.5% Firm status Government agency 14.0% 11.6% 1.6% State-owned enterprise 19.5% 31.5% 6.2% Domestic private firm 22.8% 42.6% 9.7% Foreign firm 21.9% 22.5% 4.9% Individually owned 14.2% 26.4% 3.8% Other 18.6% 28.0% 5.2% Wage (IDRs) Less than 250,000 9.8% 20.3% 4.0% 250,001–500,000 15.3% 18.3% 2.8% 500,001–1 million 20.2% 32.7% 6.6% 1–1.5 million 21.8% 52.4% 11.4% More than 1.5 million 11.9% 50.3% 6.0% Source: Sakernas (2008) survey. 23  Rather than amounts in IDRs, should we show percentages as in the WB report? Table 3 Severance pay by employee and job characteristics Legally entitled Actual Ratio of actual severance pay severance pay amount received to amount received legally entitled IDRs (million) IDRs (million) severance pay Age 15–18 years 1.274 0.145 11.4% 19–24 years 2.335 0.052 2.2% 25–34 years 7.862 1.710 21.8% 35–49 years 14.172 5.464 38.6% 50–64 years 13.201 11.727 88.8% 65 years and older 9.125 0.302 3.3% All ages 9.362 3.608 38.5% Gender Female 7.193 3.116 39.4% Male 10.184 3.886 38.2% Education Less than primary 5.136 0.773 14.3% Primary 8.148 1.894 23.3% Junior secondary 7.064 3.085 43.7% Senior secondary 11.249 5.654 50.3% Tertiary 10.577 2.204 20.8% Industry Agriculture 8.702 2.892 33.2% Mining and quarrying 35.383 14.094 39.8% Manufacturing 6.862 1.375 20.0% Utilities 8.861 0.000 0.0% Construction 7.522 0.706 9.4% 24 Trade 10.149 4.707 46.4% Transportation 14.980 6.479 43.3% Financial services 17.473 1.134 6.5% Social services 11.589 4.067 35.2% Other 10.472 5.620 53.7% Tenure Less than 1 year 0.923 0.076 8.2% 1–5 years 4.416 0.576 13.0% 5–10 years 12.662 2.155 17.0% 10+ years 22.154 13.667 61.7% Firm size 1–4 employees 7.383 2.561 34.7% 5–19 employees 6.942 1.400 20.2% 20–99 employees 8.484 1.360 16.0% 100+ employees 12.380 7.198 58.1% Firm status Government agency 4.477 0.706 15.8% State-owned enterprise 12.677 13.697 108.1% Domestic private firm 9.308 3.129 33.6% Foreign firm 9.865 5.550 56.3% Wage (IDRs) Less than 250,000 1.298 0.347 26.7% 250,001–500,000 3.843 0.724 18.8% 500,001–1 million 8.820 2.888 32.8% 1–1.5 million 14.954 5.016 33.5% More than 1.5 million 40.784 22.630 55.5% Source: IFLS (2007) survey. 25 Table 4 Severance pay by employee and job characteristics Legally entitled Actual Ratio of actual severance pay severance pay amount received to amount received legally entitled IDRs (million) IDRs (million) severance pay Age 15–18 years 1.279 0.087 6.8% 19–24 years 2.729 0.211 7.7% 25–34 years 7.958 2.527 31.8% 35–49 years 15.512 6.181 39.9% 50–64 years 21.399 2.302 10.8% 65 years and older 7.883 0.329 4.2% All ages 8.264 2.410 29.2% Gender Female 5.751 2.047 35.6% Male 9.719 2.622 27.0% Education Less than primary 8.676 2.574 29.7% Primary 5.964 1.555 26.1% Junior secondary 8.573 2.065 24.1% Senior secondary 8.477 2.931 34.6% Tertiary 13.156 5.688 43.2% Industry Agriculture 8.520 0.835 9.8% Mining and quarrying 2.200 0.000 0.0% Manufacturing 7.842 2.872 36.3% Utilities N/A N/A N/A Construction 7.101 2.873 40.5% Trade 10.928 3.544 32.4% Transportation 11.874 4.719 39.7% 26 Financial services 10.920 0.833 7.6% Public services 7.235 1.978 27.0% Tenure Less than 1 year 1.135 0.215 19.0% 1–5 years 5.024 1.018 20.3% 5–10 years 12.545 4.952 39.5% 10+ years 22.942 7.009 30.6% Firm size 1–4 employees 4.044 0.225 5.6% 5–19 employees 6.117 0.684 11.2% 20–99 employees 8.234 3.040 36.9% 100+ employees 10.431 3.283 31.5% Firm status Government agency 3.194 0.209 6.5% State-owned enterprise 21.043 5.963 28.3% Domestic private firm 9.230 3.522 38.2% Foreign firm 8.678 1.783 20.5% Individually owned 6.486 1.141 17.6% Other 6.283 1.224 19.5% Contractual arrangement Permanent 11.213 3.948 35.2% Fixed-term 8.262 1.723 20.9% No contract 7.687 2.481 32.3% Wage (IDRs) Less than 250,000 10.959 4.566 41.7% 250,001–500,000 8.552 1.129 13.2% 500,001–1 million 14.656 4.964 33.9% 1–1.5 million 13.779 4.115 29.9% More than 1.5 million 17.088 4.675 27.4% Source: Sakernas (2008) survey. 27 References Alisjahbana, Armida S. 2007. “Employment Protection Legislation and Labor Market Flexibility: International Evidence and Lessons for Indonesia – With Special Emphasis on the Cases of Outsourcing and Fixed – Term Contracts.� World Bank Office, Jakarta. Bhorat, Haroon, Ravi Kanbur, and Natasha Mayet, “Minimum Wage Violation in South Africa�, Dyson School Working Paper 2011-4 Gindling, Thomas H. and Katherine Terrel, (2007) “The Effects of Multiple Minimum Wages throughout the Labor Market: the Case of Costa Rica, Labour Economics, Vol. 14: 485-511 Holzmann, Robert, Yann Pouget, Milan Vodopivec, and Michael Weber, 2011, “Severance Pay Programs Around the World�, Policy Research Working Paper 1111 International Labour Organisation (ILO). 2007. Termination of Employment Legislation Digest.http://www.ilo.org/public/english/dialogue/ifpdial/info/termination/index.h tm ----- 2008. LABORSTA Internet. Table 2D: Total employment, by status in employment http://laborsta.ilo.org/ LP3E FE UNPAD and GIAT. 2004. “Indonesia’s employment protection legislation: Swimming against the tide.� Faculty of Economics, University of Padjajaran Bandung and Growth through Investment, Agriculture and Trade (GIAT) Project. MacIsaac, Donna. and Martin Rama. 2001. Mandatory Severance Pay: Its Coverage and Effects in Peru, Policy Research Working Paper 2626. Washington, D.C.: World Bank Development Research Group. Manning, Chris, and Kurnya Roesad, 2007, “The Manpower Law of 2003 and its implementing regulations: Genesis, key articles and potential impact�, Bulletin of Indonesian Economic Studies, vol 43 no. 1 Strauss, John, Firman Witoelar, Bondan Sikoki and Anna Marie Wattie. “The Fourth Wave of the Indonesia Family Life Survey (IFLS4): Overview and Field Report�. April 2009. WR-675/1-NIA/NICHD. Vodopivec, Milan, Lilijana Madjar, and Primoz Dolenc, 2009, “The Empirical Analysis of Severance Pay Non-Perfornance in Slovenia�, Managing Global Transitions International Research Journal, Vol. 7 no. 4. 28 Vroman, Wayne. 2007. “Reforming Social Protection for Workers in Indonesia.� World Bank Office, Jakarta. World Bank, 2011, “The Indonesian Jobs Report: Towards better jobs and security for all�, World Bank Office Jakarta.