58580 Enrollment of Firms in Social Security in Lao P.D.R. Perspectives from the Private Sector The World Bank, November 2010 Enrollment of Firms in Social Security in Lao P.D.R.: Perspectives from the Private Sector World Bank, November 2010 Summary Social Health Insurance (SHI), which is part of a broader social security system, is one of the main risk-protection schemes operating in Lao PDR and is expected to play an important role in moving the country toward universal coverage in the health sector. Although the scheme is mandatory, enrollment among the target population is low due a variety of reasons that are only partially understood. This study was conducted to better understand the factors affecting enrollment in SHI, to document employers' experiences with and perceptions of the scheme, and to identify opportunities and challenges of expanding coverage in the future. A survey was administered to 130 employers (including 65 member firms and 65 non-member firms) in Vientiane Capital. The results show that the following factors influence enrollment: state-ownership; a mixed foreign-domestic partnership; higher education level of head of company; and being in the trade industry. Explanations for not enrolling in the scheme include the lack of knowledge of the scheme, poor quality of the health care system and infrequent use of health care by employees. Member firms reported on the strengths of the scheme: reduced expenditures for the company, convenience and cost-savings for employees, and improved employee well- being. Weaknesses include: poor value for money, poor quality of care in hospitals, and a poor understanding of the social security benefits and procedures. The study revealed that, in the absence of social security benefits, non-member firms provide few benefits to their employees. Evading social security payments by hiring temporary workers does not appear to be a problem at the moment, but within firms, employees are often given the option to opt-out of the scheme. The study reveals opportunities for strengthening SHI and expanding enrollment, but also highlights the challenges of achieving high coverage rates. Expansion will require increased formalization of the labor market, effective enforcement mechanisms and effective and efficient tax collection systems. Stronger enforcement also increases incentives for evasion and informality of the labor market, while voluntary enrollment within firms makes the scheme prone to adverse selection. As the government finalizes the new health financing strategy, the costs and benefits of various approaches should be considered. However, regardless of the financing arrangement taken, increased public spending in the health sector and improved coordination across schemes can help to strengthen the health system and build up the foundation on which universal coverage can be achieved. This note was prepared as part of a World Bank program of analytic work on health financing in Lao PDR, and in collaboration with the London School of Hygiene and Tropical Medicine (LSHTM) and the Social Security Organization (SSO). Core members of the study team included Sarah Alkenbrack (LSHTM) and Magnus Lindelow (WB). The field work was implemented by Indochina Research. Helpful comments on the draft versions of the note were received from Frank Feeley, Kara Hanson, Fiona Howell, Bart Jacobs, Christoff Kurowski, Padeumphone Sonthany, Jean-Marc Thomé and Adam Wagstaff. The team would like to thank the private sector employers for participating in the survey; private sector and government staff (including hospital staff) for participating in key informant interviews; and staff members at the Social Security Organization for their assistance in providing background information. The note summarizes findings from the study; further analysis and details will be disseminated through a separate research report. Background social security scheme1 and the role it could play in expanding coverage. However, enrollment in social Increasing participation in health insurance and risk- security - which is mandatory for all enterprises with at protection schemes is a key priority for the Government of least 10 employees (according to a ministerial decree) - has Lao PDR. The four main risk-protection schemes operating remained low since the scheme's inception. Low nationally include: 1) mandatory health insurance for civil enrollment can be attributed, at least in part, to the small servants; 2) mandatory social health insurance (SHI) aimed scale of the formal sector, limited geographic reach of the at private and state-owned enterprises; 3) voluntary scheme (4 provinces only), lack of enforcement and weak community-based health insurance (CBHI) for the informal regulatory structure, and limited capacity of the Social workforce; and 4) health equity funds (HEFs) for Security Organization (SSO - the implementing agency for households living in extreme poverty. As the Government the scheme) to attract new members. of Lao PDR seeks to improve financial protection and access to health care, greater attention is being given to the 1 Although the focus of this note is on SHI, the scheme is part of a larger social security program, and therefore the terms "SHI" and "social security" are used interchangeably throughout this document. 1 basis with large companies (100 employees+) to raise Given that expansion of health insurance is a high priority to awareness about the scheme and encourage enrollment. the Government of Lao PDR, this study was conducted to However, the public relations department is staffed by only shed light on the factors affecting enrollment in social five people and lacks the financial resources and capacity to security, employers' experiences with and perceptions of effectively stimulate enrollment among the target the scheme, and benefits provided to employees in firms population [2]. that do not participate in SSO. Findings from the study will help to identify opportunities and challenges for Overview of SSO legislation defining the target group strengthening the scheme and expanding its reach, and are expected to inform decision-making related to the According to the SSO Decree 207, the official target group Government's new health financing strategy that is for social security includes all enterprises with 10 or more employees. However, the decree is difficult to enforce for currently being prepared.2 two main reasons. First, Decree 207 is a ministerial decree, and therefore by nature weaker than a law promulgated by Overview of Social Security in Lao PDR3 the President. Second, no regulatory procedures are in The social security scheme consists of a comprehensive place to enact penalties for non-compliant firms. Although package of health care and other benefits offered to an inspection unit at SSO was introduced in 2008 with a employees in private and state-owned companies. The mandate of identifying non-compliant enterprises, at the package of benefits includes medical care, sick leave, present time only verbal warnings can be issued. maternity leave, death benefits, employment injury or Furthermore, the SSO office is understaffed: like the public occupational disease benefits, retirement pensions, life relations department, the inspection unit is staffed by only 5 insurance, and disability insurance. Introduced in 1999, the employees. scheme is managed by the Social Security Organization, a semi-autonomous organization within the Ministry of Labor As a step to strengthening the regulatory framework around and Social Welfare. Funding for the scheme is generated social security, the SSO is drafting guidelines for inspections from a combination of employee and employer and sanctions for non-compliant companies, and is also contributions.4 The health care fund, which also receives working with other line ministers to promulgate a law government subsidies, is the largest of all benefits and specific to social security. The new law will mandate finances outpatient and inpatient care, and prescription enrollment of all enterprises with at least one employee.[2] drugs available at hospitals. There are no co-payments or Although the Labor Law (2006) mandates that "all labour limits on the number of contacts or services provided. units must participate in...social security", the language is vague Public providers are paid by capitation, which is fixed at in that it does not specifically define the term "labor unit" or LAK 80,000 (US$ 9.40) per insured person per year [1]. the minimum number of employees that constitute a "unit"[3]. The SSO depends on its public relations department to stimulate enrollment among its target population. The Overview of the current target group public relations team makes daily newspaper and radio While the official target group includes all enterprises with announcements, and also convenes meetings on an ad hoc at least ten (and eventually one) employees, the SSO currently targets only a small subset of enterprises in the 2 The study was designed and carried out by a team of researchers country. This is due to the way in which the SSO collects from the World Bank and London School of Hygiene and information about its target group, which involves Tropical Medicine (LSHTM). A local research team was hired to obtaining a list of businesses that are registered with the carry out data collection under the supervision of the World Bank/LSHTM team. The study was funded by the World Bank. central or provincial Tax Registration Office (Ministry of Implementation of the survey was approved by the SSO and Finance) and storing information about these businesses in a ethical approval was granted by the National Institute of database. The database includes a total of 1,357 businesses, Public Health in Laos and LSHTM in the United Kingdom. all of which are located in the four provinces where the 3 Much of the information presented here was gathered through scheme is running (Vientiane Capital, Vientiane Province, key informant interviews with stakeholders within the SSO, Savannakhet and Khammoune). The majority of this group document review and secondary data analysis. (99%) consists of firms with 10 or more employees. 4 Employees and employers contribute 4.5 per cent and 5 percent However, because the database includes only companies of employees' salaries, respectively, up to an income ceiling of that pay taxes to the Tax Registration Office, enterprises LAK 1,500,000. Health insurance makes up the largest fund, that do not have a tax identification number are excluded comprising 4.4 out of 9.5 percentage points of an employee's from the target group. earnings. 2 structure. However, even among this group, there are In order to provide some perspective of the size of the varying degrees of formality. Depending on the accounting current SSO target group relative to the universe of system, businesses make their tax payments using one of enterprises operating in the country, we performed three methods: a lump sum tax or daily tax (for firms with secondary data analysis using the Lao Economic Census [4] less advanced accounting practices), or tax paid to the Tax ­ a survey that summarizes information about all economic Registration Office on a regular basis using a tax units5 operating in the country. identification number. According to the Lao Economic Census, firms with a tax identification number account for The findings, summarized in Figure 1, are helpful for only 8% of economic units in the country (represented by understanding various options for expanding the SSO target the fourth ring). Thus, by targeting only firms with a tax group in the future. The first (outer) ring represents all identification number, the SSO automatically misses an 209,484 economic units that operated in the country at the opportunity to target the other 92% of enterprises in the time of the economic census (2006), many of which had no country. permanent address or were located in communities without road access. These units, which are usually considered part The SSO database of target firms comprises 1,320 of the informal sector6, include mobile shops, street enterprises (the fifth ring).9 However, due to low vendors, lottery sellers, tuk tuk drivers, and other compliance, only 388 enterprises were enrolled in social temporary shops. The outer ring also includes international security in January 2009 (the sixth ring). These 388 organizations and civil servants, neither of which are companies represent only 29% of companies in the SSO's considered part of the private sector, and are therefore not database of target firms, but because most enrolled firms targeted by the SSO. are large, this group accounts for 61% of targeted employees. Collectively these 388 companies provide When we exclude informal economic units, international insurance for 46,731 employees, and 86,690 organizations and civil servants, we are left with 126,913 beneficiaries10­ approximately 1.5% of the Lao economic units (represented by the second ring).7 population.[5] However, even though economic units in this category have a permanent location, the degree of formality with which Figure 1. Economic units operating in Lao PDR they operate varies. For example, only 40% of businesses in this category hold a trade registration certificate, while only 71% hold a tax registration certificate.8 In identifying opportunities to expand SSO, it is clear that targeting businesses that are not yet tax registered would present a 126,913 challenge. 209,484 economic units 90,457 economic 10,614 econ economic units with 6 ( with units 5 units tax id # permanent paying any The third ring represents the 90,457 units that are paying location) tax 4 some form of tax. Clearly, the fact that businesses are 3 paying taxes indicates that they have a somewhat formal 2 388 firms currently 1 1,320 firms reached 5 currently An economic unit is defined as any business entity, either formal targeted or informal, including shops that operate any economic activities. This group also includes non-profit and non-government Source: Compiled by the authors using data from Lao Economic organizations but excludes international organizations. Census (2005/6) and SSO Database. 6 The line between the informal and formal sector is often blurred. The informal sector usually refers to unregistered While the above diagram does not give an indication of the production units that are owned and operated by individuals, and proportion of the population that could potentially be sometimes unpaid family members. covered (see estimated scenarios presented in Appendix 2), 7 This group is likely an overestimate of the number of enterprises the data indicate that there is considerable opportunity to in the country, given that each branch office of a single company is counted as a separate economic unit. However, based on the Lao Economic Census, 99% of the economic units are solely 9 owned (not belonging to any other enterprise). The SSO target group is much smaller than the group of tax- 8 While the majority of economic units that do not have a tax registered firms, in part because the SSO targets only firms with registration certificate are very small (with less than five 10 or more employees and in 3 provinces only. 10 employees), 370 of these firms are medium-sized (10 to 99 The updated figure for number of beneficiaries at the end of employees). 2009 was 93,086 (MOH monitoring data, 2009). 3 expand the number of firms that could be reached, at least in using strata that corresponded to industry and number of principle. employees, with firms being selected from the SSO database. The survey restricted attention to Vientiane Overview of Study Capital. Key policy questions Given the objectives of the study, the questionnaire was Prior to designing the study, we conducted key informant designed to capture variables that are expected to influence interviews with government staff, leaders of business enrollment, such as characteristics of businesses associations (also known as membership organizations), and (ownership, company size, revenues, etc.), hospital staff. Through these discussions, the following key sociodemographics of company leaders, and exposure to policy questions to be addressed by the study were social security. It included both open- and closed-ended identified: questions. The survey also collected details regarding employment contracts, employee benefits, and employers' Enrollment perceptions of the scheme. Which employer characteristics are associated with enrollment in social security? (i.e. size of firm, As is often the case with firm surveys, the refusal rate was type of ownership, risk profile, etc.) ? moderately high, but this was expected given the difficulty of interviewing private sector employers.12 The firms that What motivates employers to enroll/not enroll in refused to participate differed between the enrolled and social security? unenrolled populations: in the SSO group, mostly larger Do firms intend to enroll/maintain enrollment in firms (>50 employees) refused to participate while in the social security in the future? non-enrolled group mainly smaller firms (<50 employees) refused to participate. These differences between groups Experiences may lead to some concern of sampling bias. The sample was What has been the impact of social security on weighted to adjust for the refusal rates and the firms that business? were no longer in operation or could not be found. Thus, What are the strengths and weaknesses of the the results can only draw conclusions about the SSO's social security scheme? target group of firms that are presently in business. Is the benefit package offered by non-SSO- member firms comparable to social security Findings benefits? Enrollment in Social Security Are member firms employing strategies to evade social security contributions? Background Characteristics of Sample The table in Appendix 1 outlines the relative similarities Study Approach and differences between SSO member and non-SSO To answer these questions, a structured questionnaire was member firms in the sample, with respect to a variety of administered to 130 employers (including 65 member firms background characteristics. Significantly more SSO member and 65 non-member firms), interviewing heads of firms are state-owned, under foreign ownership, and have a companies where possible.11 The sample was randomly larger workforce and higher revenues. SSO firms also selected from the four biggest industries in the SSO target employ fewer permanent workers (relative to the size of group: manufacturing; construction; trade; and services, their workforce) than non-member firms. However, it is 12 The response rate was between 67% and 80%. The former rate 11 Pilot-testing took place in February 2009. Field work was includes contacted firms in which nobody was available to be conducted by a local research team from March through May interviewed or a follow-up appointment could not be made in the 2009. Thirty-nine percent of interviews were conducted with the study period. The latter rate includes only firms that explicitly head of the company (e.g. director, general manager, chief refused to participate. Interviewers attempted to schedule an executive officer, etc.); 30% were conducted with the head interview three times before randomly selecting another accountant; and 33% were conducted with the human resources enterprise. It is important to note that one third of the firms that manager. (Some were conducted with more than one were identified as eligible had either closed or could not be found respondent). Given that the study focused on benefits and (e.g. the number was disconnected or the business was no longer contracts it was reasonable for the head accountant or human at the stated location), indicating the poor quality of the database resources director to participate when the head of the company used by SSO. (These firms are not factored into the response was not available. rate.) 4 important to note that differences noted here are simply represents a lower likelihood of enrolling, relative to the differences in the sample: further analysis using multivariate reference group. Given the small sample size, we consider a logistic regression allows us to examine differences in key p-value of less than 10% to be significant. The findings variables while holding all other factors constant. These show that companies whose heads have a university findings are presented as odds ratios below. education are more than three times more likely to enroll in social security, over companies whose leaders have less than a university education. The odds of enrolling are also Which employer characteristics are associated with enrollment significantly higher for firms in the trade industry relative to in social security? the services industry. The odds ratios presented in Table 1 compare the odds of enrolling in social security for companies with and without Ownership was found to be associated with enrollment. certain characteristics. The odds ratios represent the The odds of enrolling are approximately 23 times higher for strength of the relationship between the variable of interest state-owned enterprises than for private firms, which is not and enrollment in social security when all other variables surprising given that social security is a government- are held constant.13 An odds ratio greater than one mandated program.14 Although foreign-owned companies represents a greater likelihood of enrolling, relative to the are not significantly more likely to enroll, mixed companies Table 1. Odds ratios of enrollment, by firm characteristic Odds Ratio 95% Confidence p Interval Head of company has a university education 3.475 1.188 10.162 * Head of company is male 0.627 0.177 2.217 - Industry type (Reference group: services) Manufacturing 1.210 0.289 5.056 - Construction 0.320 0.038 2.698 - Trade 3.697 0.845 16.171 (*) Ownership 1: State-owned or partially state-owned (ref: 100% private) 23.134 1.987 269.364 * Ownership 2: (Reference group: 100% domestic) 100% foreign (vs. domestic) 6.174 0.319 119.270 - mixed (vs. domestic) 20.282 2.965 138.751 * More than one service outlet/branch/factor/store 1.505 0.421 5.375 - Revenues (Reference group: <1 billion kip) 1-10 billion Kip 1.342 0.443 4.071 - > 10 billion Kip 1.300 0.299 5.657 - Member of business organization (ref: no membership) 2.064 0.728 5.857 - Firm has higher than avg perceived risk (3+ on a scale of 1-5) 0.707 0.102 4.910 - Size of permanent workforce =20-59 employees (ref: 0-19 emp) 1.998 0.568 7.026 - Size of permanent workforce = 60+ employees (ref: 0-19 emp) 3.550 0.815 15.456 (*) Employs any temporary workers 1.202 0.388 3.726 - At least 10% employee turnover (annually) 0.522 0.159 1.711 - Taxes paid at central level (ref: taxes paid at lower levels) 0.675 0.160 2.848 - *significant at the 5% level (*) significant at the 10% level. A p-value is a statistical term that measures the likelihood that a difference observed in the sample is due to chance. A result that is significant at the 5% level has a p-value of less than .05. The confidence interval indicates the reliability of the estimated odds ratio. A 95% confidence interval means that there is a 95% chance that the true estimate falls within the interval. Because the sample is small, p-values that are less than .10 are accepted as (borderline) significant. reference group, while an odds ratio of less than one 14 In this sample, all but one state-owned or partially state-owned company are enrolled in social security. Therefore, the effect of state-ownership on enrollment is very strong. It is not clear whether there is an additional mechanism for SOEs to enroll in 13 Given the cross-sectional nature of the data, the associations social security. For example, it is possible that SOEs are pressured to enroll or are automatically enrolled. Whether or not observed here do not imply causal relationships. this is the case will require further investigation. 5 (in which ownership is shared between domestic and through by the SSO. However, it is more likely that in their foreign owners) are 20 times more likely to enroll in social explanations of why they were not enrolled, employers security than domestic companies. It is possible that mixed tried to justify their behavior in terms that did not make companies, due to the nature of their business, receive them look bad. Other important reasons for not enrolling pressure to comply with industry regulations, or that these were: employers' lack of knowledge of social security, the types of firms have a stronger compliance culture. poor quality of government hospitals, and the fact that employees do not use benefits or do not get sick. Prior to the interviews, key informants reported that they expected enrollment to be higher for larger firms. The To better understand the process by which a firm decides to study confirmed that larger firms are more likely to enroll, enroll in social security, employers were asked to identify all other factors being equal. However, a firm's decision to the people in charge of making the decision to enroll. Most enroll is not significantly influenced by a firm's perceived firms mentioned that the director of the company, the risk, employee turnover, or financial status. Although key general manager or the owner, and less often the executive informants expected pressure from leaders of membership board, make the decision to enroll. In a few companies, the organizations to influence employers to enroll in SSO, this personnel manager and all staff members are involved in the claim was not supported by the study findings. decision process. What motivates employers to enroll/not enroll in social security? Table 2. Most important reasons for enrollment/ non-enrollment To better understand employers' decisions to enroll, Reasons for enrollment Rating (1-5) employers were asked to rate a number of possible reasons To ensure employees have health care 4.28 for enrollment/ non-enrollment on a scale of 1 to 5, with 1 coverage being not at all important and 5 being very important.15 The To increase employee satisfaction 4.11 To improve health and well-being of 4.08 results, summarized in Table 2, confirm the trends employees identified through multivariate analysis but also shed further To ensure employees have retirement benefits 4.00 light on employers' decision-making process. Among the Strong pressure from international bodies 3.05 insured, the most important reason for enrollment was to Strong pressure from SSO 2.74 ensure employees are covered with health insurance. Strong pressure from employees 2.72 Increasing employee satisfaction, and improving the health Strong pressure from membership 2.71 and well-being of employees were also important factors. organization These two findings indicate that employers are most Reasons for non-enrollment Rating (1-5) concerned with the health insurance benefits within the Company benefits are better than SSO benefits 4.11 social security scheme. However, retirement benefits are Do not know much about social security 3.42 also relatively important.16 The finding that pressure from Quality of government hospitals not good 3.22 the SSO was not an important factor was not surprising Do not use health care benefits/ staff do not 3.22 get sick given that SSO is not currently enforcing enrollment. Employees do not want SSO 3.12 Cost of SSO is too high 2.97 Among the non-enrolled cohort of the sample, the most Do not trust that money is used well 2.83 important reason for non-enrollment was that the High turnover among employees 2.57 company's benefit package is superior to social security ­ Many temporary employees 2.38 Employees prefer to purchase private 2.35 an interesting finding given that few companies offer health insurance insurance or other benefits to their employees, as shown in the next section of the report. Therefore the only possible explanation of this finding is that although non-enrolled Do firms intend to enroll/maintain enrollment in SSO in the firms offer fewer benefits to their employees, the benefits future? they do offer are of higher quality than those offered The study found that almost all SSO member firms (95%) 15 The possible factors affecting enrollment/non-enrollment were intend to be enrolled in the scheme in 3 years, compared identified by first asking employers to list their most important with 38% of non-member firms (51% are unsure and 11% reasons for enrolling/not enrolling. Respondents were then asked said they do not intend to be enrolled). The majority of to rank the full list of responses. firms that are "unsure" about enrollment or do not intend 16 In a survey question, respondents ranked health care as the to enroll are domestic, privately owned businesses. most important benefit, followed by retirement benefits and sick Promotion activities targeted at leaders of this type of firm leave. 6 ­ e.g. face-to-face meetings to discuss SSO and answer Strengths questions about the scheme ­ could help encourage Social security reduces employers' expenditures on benefits enrollment. Currently, only 13% of non-enrolled firms in for which the company would otherwise pay. the sample have ever attended a meeting where SSO was promoted. Social security provides convenience and cost-savings to employees, in that employees can use cards for hospital care, rather than paying cash. Experiences with social security Being enrolled in social security improves employee well- being. Employers report that SSO gives employees a What has been the impact of social security on business? sense of "security", "warm feelings", "trust", and As shown in Figure 2, the majority of respondents (69%) "confidence" and "increases employee satisfaction". stated that social security has had a positive effect on According to one employer, higher employee employee well-being and satisfaction. However, most feel satisfaction due to SSO benefits translates into higher that the scheme has had no effect on the ability to retain productivity. staff or recruit staff members. Employers explained that Weaknesses while some prospective employees inquire about benefits prior to accepting a position, most job-seekers are mainly Employers do not perceive the scheme to be of good value, concerned with salary. either because the staff do not use benefits (most respondents made specific reference to not using health Although 54% of SSO-enrolled employers claim that the care benefits) or because contributions exceed the impact on overall business has been positive, almost half say benefits used. Employers are particularly concerned there has been no effect. Among firms reporting a positive that health insurance does not cover check-ups and feel effect, employers explained that social security: a) gives that including free check-ups would improve the value firms a competitive advantage over non-enrolled firms of the scheme. because it decreases company expenditures; b) helps to Quality of care at hospitals is poor, according to SSO build trust with staff members; and c) increases members - and worse for social security members. For productivity of employees. However, other employers example, SSO employers report that their employers noted that being enrolled in social security decreases experience longer waiting times, are treated poorly by revenues unnecessarily because the cost of contributions health care professionals, and receive drugs of inferior exceeds the benefits used by employees. quality, relative to non-members. One respondent mentioned that because of the long waiting times and Figure 2. Impact of Social Security on Business poor quality "...members sometimes have to go to the private 100% clinic for treatment." 80% Communication by SSO regarding benefit structure and the procedures associated with the scheme is poor. Members 60% reportedly have very little contact with SSO staff and 40% therefore no opportunity to learn more about SHI. 20% Some SSO representatives do not clearly explain the details of the scheme. Moreover, there are reports that 0% SSO staff members are unfriendly when providing employee well- ability to retain or overall business information to members or prospective members. being and recruit staff satisfaction Other less frequently reported complaints of the scheme Positive effect No effect Negative effect Not sure include slow and cumbersome documentation processes required to submit health care claims and the difficulty of What are the strengths and weaknesses of the social security using (or switching) hospitals for the primary source of scheme? treatment. Employers who are currently enrolled in social security were asked to comment on the strengths and weaknesses of the scheme. The three most common responses are described below. 7 Is the benefit package offered by non-member firms comparable to of time an employee can work as a temporary employee social security benefits? varies across firms, and 30% of firms report that there is no limit. Because the majority of temporary employees do not As stated earlier, the most important reason for not receive benefits, there may be a financial incentive for enrolling in social security is that employers offer a better employers to hire temporary employees to fulfill the duties package of benefits than that offered by the SSO. However, of permanent employees. The potential for such evasion is the study findings show that the non-enrolled firms offer far discussed in the next section. fewer benefits to their employees than do SSO member firms (See Figure 3). In fact, the majority of non-enrolled Are member firms employing strategies to evade social security employers do not offer health care benefits: only one contributions? quarter of non-SSO firms make direct payments for employees' health care, while none have private insurance. One of the challenges of implementing and enforcing In contrast, some SSO member firms offer extra insurance mandatory insurance is evasion of contributions. For for programs that are already included in the social security example, the international body of literature on social package (e.g. private health insurance, maternity benefits, security describes the ways in which many enterprises fail coverage for injuries, and sick leave). to comply with social security (e.g. by underreporting the number of permanent employees, reclassifying job Although private insurance is not commonly offered at the descriptions - in order to hire temporary workers instead of firm level, it is possible that employees in non-SSO member permanent, underreporting earnings, and failing to remit firms are purchasing private insurance at the household employee contributions) [7-9]. In Lao PDR, it was level. Key informant discussions found that AGL, and to a expected that any evasion of social security payments would lesser extent, Toko Assurance, are the two players in the be minimal, given that penalties are not yet enacted for private insurance market in Lao PDR and plans are non-compliance. underway for rapid expansion. AGL has launched new initiatives, including cooperative arrangements with One way to detect invasion is to examine the structure of International NGOs and villager groups, whereby the workforce. A disproportionately high number of premiums are financed by income generating activities. temporary workers could indicate that a firm is trying to AGL offers a benefit package for as low as US$ 2.50 per evade social security contributions. The study found that month and covers care both within and outside Lao PDR overall, enrolled firms employ more temporary and daily [6]. workers than non-enrolled firms (24% vs. 15%; the differences are significant at the 10% level, as shown in Figure 3. Benefits offered to SSO and non-SSO employees Appendix 1). However, when we control for type of industry (See Figure 4), the difference in the proportion of 100% non-permanent workers between SSO and non-SSO firms 80% SSO non-SSO is very small and is not statistically significant. Furthermore, when controlling for differences in size of 60% workforce, revenue, and ownership (private vs. state- 40% owned; foreign vs. domestic), the differences almost 20% disappear. It seems that SSO firms are no more likely to 0% employ temporary workers than non-SSO member firms ________________... and that the extent to which firms hire temporary workers Direct payment of HC Paid annual leave Direct payment of HC Paid annual leave Social security Paid maternity leave Paid sick leave Social security Paid sick leave Private health ins. Private health ins. Injuries/disability is likely related to the nature of the business. For example, construction companies tend to hire workers on a seasonal basis, explaining the higher proportion of temporary workers in the construction industry. Permanent employees Temporary employees With respect to temporary employees, very few in either group receive employee benefits. However, 15% of SSO firms do extend social security and pay annual leave to their temporary employees. According to employers, the length 8 Figure 4. Composition of workforce by industry and SSO status permanent workers in the company with the number of employees enrolled in social security for the same time 100% period. If firms are complying with social security, 100% of 80% the permanent workforce should be enrolled. However, 60% the findings presented in Figure 5 show that 44% of 95% member firms enrolled less than 100% of the permanent 40% 73% 76% 82% 91% 81% workforce in the scheme. In fact, 94% of employers with 20% 52% 47% social security benefits openly admitted to giving employees 0% the option to enroll. Voluntary enrollment within firms poses a challenge to expansion of social security because it decreases revenues to the SSO. Voluntary enrollment also encourages healthy and lower-risk individuals to leave the risk pool. This phenomenon of adverse selection could Daily Temporary Permanent ultimately threaten sustainability and effectiveness of risk- pooling. Given that enrollment is not enforced, it makes sense that member firms are no more likely to evade payments than Figure 5. Percent of permanent workforce enrolled with SSO are non-member firms. However, it is possible that, regardless of enrollment, firms are evading payment of 60 % of SSO firms in sample benefits in general. The language in the Labor Law is such 50 that even employers without social security are required to 40 provide sick leave, maternity benefits, health care, occupational disease benefits, and pensions. These expenses 30 must be borne by the employer for firms that do not have 20 social security. Therefore, it is possible that firms (both 10 SSO and non-SSO members) hire temporary workers, or 0 under-report number of permanent employees as a way of 100+ 100 <10 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-99 avoiding these payments. For example, key informant discussions with the SSO revealed that after enrolling with Percent of permanent workforce enrolled the SSO, some companies reported gradual declines in the size of their permanent workforce. However, according to *Firms that offer SS benefits to both permanent and temporary SSO staff, random checks by their organization revealed employees are effectively enrolling more than 100% of their that in some of these firms the size of the permanent permanent workforce. workforce had not actually decreased but that the firm had instead underreported the number of employees [2]. Underreporting the number of employees is a well-known Looking forward: Challenges and Opportunities evasion tactic for reducing social security payments [8], and This study brings to light a greater understanding of may become a growing concern in Lao PDR if SSO employers' attitudes toward social security. Although there strengthens enforcement. Stronger regulation could create are limitations to the study17, the results are nonetheless incentives for employers to not only evade social security helpful for understanding perspectives of the private sector payments by under-reporting employees but could also lead ­ a group that is not usually engaged in health sector reform to a growth in the informal sector (because firms wishing to discussions. A number of issues also emerge from the study, evade contributions will fail to register some or all of its which point to the following operational changes that could workers). A rise in informality also decreases the tax base strengthen the scheme and facilitate expansion: for general revenues [8, 10]. Prior to the study, key informant interviews suggested that some employers allow employees to opt-out of social 17 security. Given that social security is mandatory for all The narrow sampling frame limits the generalizability of the permanent employees, opt-outs can be considered an findings; the high refusal rate and the use of different respondents evasion strategy used to reduce costs of providing benefits. (e.g. head of company, HR director) may bias the responses; the small sample leads to large variation in estimates and large To further investigate the extent to which opt-outs take confidence intervals, which can increase the likelihood of place, researchers compared data on the number of sampling errors. 9 Improve awareness-raising, promotion and communication. himself than making contributions to the social security Awareness-raising activities and face-to-face meetings scheme. For those with more than 20 years left until with employers from privately-owned domestic firms retirement, it makes sense to contribute only for 30 ­ the least likely to enroll ­ could be effective years in total.[12] The results from Vietnam indicate strategies for bolstering enrollment rates. However, that reforming pensions, and possibly other benefits, new approaches will be required to reach smaller firms will make social security more attractive to private with less formal structures. For example, large hotels sector employers and may facilitate the scheme's that have been enrolled in social security for several expansion. The same could be true for Lao PDR, years have begun to realize the benefits to their bottom although to be sure, further research along the lines of line (i.e. because they do not have to cover direct the Vietnam analysis could be undertaken. payments of health care and sick leave) and are now working with the Lao Hotel and Restaurant Association In addition to the findings generated from the study to encourage smaller guest houses to join the scheme directly, the descriptive analysis of the social security [11]. Similar partnerships may be feasible in other scheme (in the background section of this report) highlights sectors. several challenges to the scheme. With these challenges in mind, the following operational changes are proposed: Employers of SSO-member firms complain about the content, frequency and style of communication by SSO Introduce legislation to enforce enrollment. The scheme is (e.g. SSO members reported that SSO staff are not effectively voluntary at present (for firms, and also for well-informed about the scheme and also treat SSO employees within firms). Although inspections have members poorly). Strengthening communication skills begun, the SSO will be in no position to penalize non- and increasing communication channels between the compliant firms until a law is enacted. Completion of SSO and businesses could therefore help to improve the guidelines for inspections, which will be drafted in satisfaction among members, and could ultimately the near future, will also be important for attract new members. strengthening enforcement but without a law, inspections and sanctions will carry little weight. Promote improvement in quality of health services. Poor quality of care in government hospitals is a major Build capacity of SSO office. Efforts to increase reason that employers do not enroll in social security. promotion or strengthen enforcement will require Thus, quality improvements to the system as a whole expansion of various departments (e.g. the inspections would have a positive impact on enrollment. Although department; the public relations department) and quality improvement is a complex issue requiring capacity building. Given the poor image of SSO, public multifaceted interventions, there is nonetheless information and customer service training will be potential for the SSO to influence service delivery by critical for SSO. However, it may be beneficial to using its purchasing power to encourage better quality contract-in certain tasks such as the development of care. For example, the SSO could make a percentage public information strategies and independent of its capitation payment contingent on quality of customer feedback surveys. services, or issue penalties to staff members who treat patients poorly. Expand target group. As the SSO looks to increase coverage of social security in the future, there are Make the social security package, as a whole, more attractive. several possible routes it could consider. These Although SHI is the largest and most important benefit include: 1) expansion within the current target group; within the social security package, it is important to 2) expansion to new geographic areas; 3) expansion to note that enrollment in SHI is linked to the broader smaller enterprises; and 4) expansion of the target package of social security benefits and that decisions to group beyond the narrow group of tax-paying firms. enroll in health insurance are therefore determined by To get an estimate of the population that could the perception of the benefit package as a whole [12]. realistically be reached, scenarios were developed using For example, if retirement benefits (the second most various assumptions. These scenarios are described in important benefit in the scheme) are not financially detail in Appendix 2. The second scenario shows that attractive to workers, employers may be discouraged even if compliance among the current target from enrolling in social security. A study in Vietnam population increases to 80% of beneficiaries, increased found that a private sector worker with less than 20 enrollment would result in a coverage rate that ­ on a years left before retirement is better off saving for population level ­ is not substantially different from 10 the current coverage rate. If the target group were challenges may be more difficult to overcome. First, there expanded to "hard-to-reach" groups (e.g. firms with are high costs to stronger enforcement, as offices will need less than 10 employees, firms without a tax to be established throughout the country (currently only identification number, firms not paying any tax), one national office and two provincial offices exist) and coverage rates would still remain relatively low infrastructure and legislation required for enforcement will because because compliance for less formal and smaller need to be developed. The costs would likely increase firms would be difficult to enforce, and because considerably once smaller and less formal firms are smaller firms do not add substantially more employees. targeted. Furthermore, stricter enforcement can exacerbate Although the cost estimates of the scenarios are not incentives for informality and evasion, the costs of which included, it is likely that expansion of the target group can be significent [10]. For instance, in Colombia, evasion to these "hard-to-reach" groups will result in of social security payments among formal sector workers diminishing returns. Moreover, any increase in was estimated to cost US$836 million in forgone revenues coverage of the current scheme is likely to reach the in a year (2.75% of GDP),[14] while in the Philippines and better off, and is unlikely to help the country achieve Kazakhstan, 30% and 40% of expected revenues, broader goals of improved equity and poverty respectively, were actually collected[15, 16]. reduction. Apart from the costs relating to enforcement and evasion, Improve tax registration. Given the small number of SHI also has other implications that need to be considered. firms with a tax identification number (8%), Adverse selection ­ which is present in the Lao scheme ­ monitoring tax payments will remain a challenge, even can lead to cost escalation by increasing the average cost of if enforcement is enhanced. One way to address this health care per member. The scheme may be forced to raise challenge would be to bring more firms into the formal contributions to cover the costs, leading to further evasion, tax system ­ e.g. by linking tax and business or look to general tax revenues to cover the deficit. registration more clearly. Furthermore, as noted in the Strengthening enforcement within and across firms could World Bank's survey of "Doing Business in Laos", there help to reduce adverse selection but would not eliminate it. are various bottlenecks of the business registration A universal coverage scheme on the other hand, would process that could be simplified, thereby making it eliminate adverse selection by enrolling the entire easier to monitor payments [13], population, but would of course bring other challenges [12]. Consolidate administration to enhance administrative efficiency. In the long run it may be more efficient to From the synthesis of the issues highlighted by this study, it collect and enforce SHI premiums and other is clear that although the social security system can play an contributions through a single agency (e.g. the tax important social insurance role for the small but growing office, Ministry of Finance), rather than through the formal sector, it is unlikely to be a vehicle for improving SSO. This approach has been used in Argentina and is access to health services, reducing health-related financial expected to facilitate compliance because employers risk on a broad basis, or generating substantial revenues for and workers need make payments only to a single the health sector. To achieve these goals, complementary agency rather than to multiple agencies [8]. However, approaches will be needed. There is scope for building on the advantages and disadvantages of combining other health protection schemes, including community- collection in Lao PDR would have to be analyzed based health insurance and health equity funds. Currently, further. the different health protection schemes are highly fragmented. However, the new health financing strategy Establishing new reporting procedures could also help envisages that the schemes will be merged over the next to minimize evasion. For example, in Ukraine, a new few years [17]. This would be an important step towards reporting procedure requires employers to indicate the improved coordination and reduced administrative costs. names of employees for whom the social security Integration under one organization could also increase contributions are being made. This system helps to purchasing power and facilitate better supervision of identify employers that do not contribute for each quality. However, given the current low coverage rates it is employee[8]. However, a certain level of sophistication unlikely that simply merging the schemes will facilitate in enforcement is required. expansion of enrollment. But while these operational changes can strengthen SHI and To address the challenge of low coverage, substantial expand enrollment among the target population, other government (or donor) financing will be required. 11 Currently, government spending on health in Lao PDR is 9. Wagstaff, A., Social Health Insurance vs. Tax- very low by international standards, but fiscal space is set to Financed Health Systems - Evidence from the OECD, in increase in coming years due to growing natural resource Policy Research Working Paper #4821. January 2009, revenues and improved tax administration. Insofar as the The World Bank: Washington, DC. government opts to use this space to increase spending on 10. Wagstaff, A., Social health insurance reexamined. health, these funds could be used to either subsidize health Health Econ, 2010. 19(5): p. 503-17. insurance coverage directly, or increase funding to health 11. Personal communication with Lao Hotel and care providers (potentially based on volume or other Restaurant Association. June 2008. measures of performance) and reduce the charges for 12. Lieberman, S.S. and A. Wagstaff, Health Financing patients. Both options can contribute to improved access and Delivery in Vietnam: Looking Forward. 2009, and financial protection, and both have advantages and Washington, DC: The International Bank for disadvantages. These issues need to be studied and Reconstruction and Development/ The World discussed in more detail and the benefits and trade-offs of Bank. 171. achieving universal coverage through different modes of 13. The World Bank, Doing Business 2010: Lao PDR. financing need to be reviewed. 2009, The World Bank and International Finance Corporation. Although the recommendations from this study do not 14. Escobar, M.-L. and P. Panopoulou, Chapter 22. provide a prescriptive plan for the way forward, they do Health, in Colombia: The Economic Foundation of provide a platform for dialogue during the development of Peace, M. Giugale, O. Lafourcade, and C. Luff, the health financing strategy. As the government sets forth a Editors. 2003, World Bank: Washington, DC. p. strategy for health financing reform, it will be critical to 653-707. base plans for the future on a realistic assessment of what 15. Jowett, M. and W.C. Hsiao, The Philippines: can be achieved through current health protection schemes, Extending Coverage beyond the Formal Sector, in Social and to be clear on the government expenditure Health Insurance for Developing Nations, W. Hsiao commitments that are needed to achieve health sector and R.P. Shaw, Editors. 2007, The World Bank: objectives. Washington, DC. p. 81-104. 16. Gottret, P. and G. Schieber, Health Financing Revisited: A Practitioner's Guide. 2006, Washington, References DC: World Bank. 17. MOH Lao PDR (Department of Planning and 1. Thomé, J.-M. and S. Pholsena, Chapter III: Lao Finance), Draft Health Financing Strategy, Lao PDR, People's Democratic Republic: Health Financing Reform 2011-2015: Vientiane, Laos. and Challenges in Expanding the Current Social Protection Schemes, in Promoting Sustainable Strategies to Improve Access to Health Care in the Asian and Pacific Region. 2009, UNESCAP. 2. Personal communication with Social Security Office. December 17, 2008: Vientiane, Laos. 3. President of the Lao People's Democratic Republic, Labour Law, P.s. office, Editor. 2006, National Assembly: Vientiane. 4. National Statistics Center, Lao Economic Census. 2005/2006. 5. Social Security Office, Database on enrolled firms. 2008. 6. Personal communication with AGL and Toko. 2009. 7. McGillivray, W., Contribution evasion: Implications for social security pension schemes. International Social Security Review, 2001. 54(4): p. 3. 8. Bailey, C. and J. Turner, Strategies to Reduce Contribution Evasion in Social Security Financing. World Development, 2001. 29(2): p. 385-393. 12 Appendix 1. Background characteristics of sample Firm Characteristics SSO (n=65) Non-SSO (n=65) Difference Industry category Manufacturing 30.8% 33.8% Construction 9.2% 6.2% n/a Trade 20% 24.6% Services 40% 35.4% 100% 100% Ownership (private/state) Ownership 1: State-owned or partially state-owned 26.2% 1.5% 0.001* (0=100% private) Ownership 2: 100% domestic 39(60%) 54(83.1%) 100% foreign 14(21.5%) 9(13.9%) 0.005* Mixed partnership (domestic & foreign) 12(18.5%) 2(3.1%) Total 65(100%) 65(100%) Employees Mean # of permanent employees 80.5 63 0.461 Mean # of temp employees 36.9 7.5 0.024* Mean # of daily wage employees 4.6 0.9 0.178 Permanent employees as a % of workforce 75.8% 84.5% 0.094** % of workforce female 37.4% 41.1% 0.408 Size of company (financial) Company Assets < 1 billion kip 27.7% 40.0% 1-10 billion kip 41.5% 43.1% 0.130 > 10 billion kip 30.8% 16.9% Company Revenues < 1 billion kip 29.2% 49.2% 1-10 billion kip 41.5% 40% 0.012* > 10 billion kip 29.2% 10.8% Other firm characteristics >1 service outlet/factor/branch/store 78.5% 76.9% 1.000 Member of business organization 56.9% 36.9% 0.035* Higher than average risk 16.9% 13.9% 0.627 Age distribution (at least 50% of employees are <35 years) 87.7% 96.9% 0.096** Mean company turnover (annual) 10.0% 10.1% 0.972 Level of tax payment central 41.5% 20.0% provincial 47.7% 50.8% 0.005* district 10.8% 29.2% Characteristics of company head Nationality (head is Laotian) 69.2% 83.1% 0.099** Mean age (years) 52.6 49.8 0.131 Gender (head is male) 83.1% 75.4% 0.387 Education of head (attended university/college or 83.1% 55.4% 0.001* higher) *significant at the 5% level **significant at the 10% level. A p-value is a statistical term that measures the likelihood that a difference observed in the sample is due to chance. A result that is significant at the 5% level has a p-value of less than .05. This also means that there is a 95% chance that the value is not due to chance. 13 Appendix 2. Scenarios for Increasing Coverage The scenarios depicted in Figure 6 were constructed using Given that most of the firms in areas that are not currently data from the Lao Economic Census, and assumptions about covered are small and medium-sized firms, compliance of the feasibility of reaching each target group. The scenarios scenario 3 is expected to decrease, relative to the are based on assumptions about future compliance, and compliance level in scenario 2. or illustrative purposes, we neither account for population growth nor expansion of the assume compliance will decrease to 70%. Scenario 4 targets target groups. Nor do they account for the impact of firms with at least 10 employees who pay any kind of tax. increased compliance on evasion or employment. Relative to scenario 3, this scenario adds an additional Validation of scenarios will require more precise data as 83,000 beneficiaries and although compliance is expected well as consensus building with key decision-makers about to be reduced (to 65%), it may be more cost-effective to the target group and expected compliance. Although the target firms that pay any tax, than to only target firms with scenarios are simplistic approximations, they highlight the tax identification numbers. However, identifying these opportunities and challenges associated with expansion of firms may be difficult if the firms have not carried out the target group. formal licensing requirements. Beginning with the baseline scenario (scenario 1), we see The next 3 scenarios represent expansion to groups that are that current compliance stands at 29% of firms but covers increasingly more difficult to reach. Scenario 5 includes 61% of employees. The higher coverage among the latter is firms with one or more employee with a tax identification a result of covering mainly large enterprises. Given that the number, and compliance is expected to be the same as in infrastructure required to strengthen enforcement is almost scenario 4 (65%). In Scenario 6, firms of any size and non-existent in Laos, considerable effort will be required to paying any kind of tax would be targeted. We assume increase compliance. Nevertheless, scenario 2 assumes compliance would fall to 60%, given the difficulty of compliance within the current target group reaches 80%. reaching these smaller, informal firms. Finally, Scenario 7 Given the small size of the SSO target group relative to the includes firms of all sizes, regardless of their tax payment. universe of firms in the country, as shown earlier in Figure We assume that compliance decreases to 55% for this 1, increasing compliance to 80% will not result in group. As each new "hard to-reach" group is added, substantial increases in coverage from a population compliance is expected to decrease. Thus, the cost of perspective. enforcing enrollment among harder-to-reach groups, such as smaller enterprises and enterprises with less formal tax The remaining scenarios illustrate various options for arrangements, will likely result in diminishing returns in gradually expanding the target group. Scenarios 3 through 7 that the premiums collected from these additional groups represent target groups that include all accessible areas will be less than the cost of identifying these firms and throughout the country. In scenario 3, the SSO would enforcing enrolment. continue to target firms with 10 or more employees, and only firms with a tax identification number. Figure 6. Scenarios for expanding coverage of social security 1,200,000 1,000,000 800,000 600,000 400,000 200,000 0 1. Baseline 2. Current target 3. Geographic 4. Geographic 5. Expanded scope 6. Expanded scope 7. All firms with 1 scenario group with 80% expansion (10+ expansion (10+ (1 + employees, (1+ employees, + employees, any compliance employees, tax employees, any tax id) any tax) tax and no tax id) tax) Expected coverage (med compliance) # Beneficiaries targeted 14