OIT - Compliance with Labour Regulations in West African Countries A Multi-Dimensional and Dynamic Analysis

X Compliance with Labour Regulations in West African Countries A Multi-Dimensional and Dynamic Analysis Authors / Thierno Malick Diallo, Lucas Ronconi

October / 2025 ILO Working Paper 152© International Labour Organization 2025 Attribution 4.0 International (CC BY 4.0) This work is licensed under the Creative Commons Attribution 4.0 International. See: https:// creativecommons.org/licenses/by/4.0/. The user is allowed to reuse, share (copy and redistribute), adapt (remix, transform and build upon the original work) as detailed in the licence. The user must clearly credit the ILO as the source of the material and indicate if changes were made to the original content. Use of the emblem, name and logo of the ILO is not permitted in connection with translations, adaptations or other derivative works. Attribution – The user must indicate if changes were made and must cite the work as follows: Diallo, T., Ronconi, L. Compliance with Labour Regulations in West African Countries: A Multi-Dimensional and Dynamic Analysis. ILO Working Paper 152. Geneva: International Labour Office, 2025.© ILO. Translations – In case of a translation of this work, the following disclaimer must be added along with the attribution: This is a translation of a copyrighted work of the International Labour Organization (ILO). This translation has not been prepared, reviewed or endorsed by the ILO and should not be considered an official ILO translation. The ILO disclaims all responsibility for its content and accuracy. Responsibility rests solely with the author(s) of the translation. Adaptations – In case of an adaptation of this work, the following disclaimer must be added along with the attribution: This is an adaptation of a copyrighted work of the International Labour Organization (ILO). This adaptation has not been prepared, reviewed or endorsed by the ILO and should

not be considered an official ILO adaptation. The ILO disclaims all responsibility for its content and accuracy. Responsibility rests solely with the author(s) of the adaptation. Third-party materials – This Creative Commons licence does not apply to non-ILO copyright materials included in this publication. If the material is attributed to a third party, the user of such material is solely responsible for clearing the rights with the rights holder and for any claims of infringement. Any dispute arising under this licence that cannot be settled amicably shall be referred to arbitration in accordance with the Arbitration Rules of the United Nations Commission on International Trade Law (UNCITRAL). The parties shall be bound by any arbitration award rendered as a result of such arbitration as the final adjudication of such a dispute. For details on rights and licensing, contact: rights@ilo.org. For details on ILO publications and digital products, visit: www.ilo.org/publns.

ISBN 9789220427699 (print), ISBN 9789220427705 (web PDF), ISBN 9789220427712 (epub), ISBN 9789220427729 (html). ISSN 2708-3438 (print), ISSN 2708-3446 (digital) https://doi.org/10.54394/EBGO7731

The designations employed in ILO publications, which are in conformity with United Nations practice, and the presentation of material therein do not imply the expression of any opinion whatsoever on the part of the ILO concerning the legal status of any country, area or territoryor of its authorities, or concerning the delimitation of its frontiers or boundaries. See: www.ilo. org/disclaimer. The opinions and views expressed in this publication are those of the author(s) and do not necessarily reflect the opinions, views or policies of the ILO. Reference to names of firms and commercial products and processes does not imply their endorsement by the ILO, and any failure to mention a particular firm, commercial product or process is not a sign of disapproval. Information on ILO publications and digital products can be found at: www.ilo.org/researchand-publications ILO Working Papers summarize the results of ILO research in progress, and seek to stimulate discussion of a range of issues related to the world of work. Comments on this ILO Working Paper are welcome and can be sent to research@ilo.org.

Authorization for publication: Caroline Fredrickson, Director, Research Department ILO Working Papers can be found at: www.ilo.org/research-and-publications/working-papers Suggested citation: Diallo, T., Ronconi, L. 2025. Compliance with Labour Regulations in West African Countries: A Multi-Dimensional and Dynamic Analysis, ILO Working Paper 152 (Geneva, ILO). https://doi.

org/10.54394/EBGO773101 ILO Working Paper 152 Abstract This paper provides measures of compliance with labour regulations that vary across space, time, and type of regulation, allowing for a rich analysis of worker vulnerability. We compute a multidimensional index of labour violations that covers minimum wages, leave benefits and social security in eight West African countries (Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, Togo), and measure the persistence of labour violations experienced by employees over time (2019 to 2022). Overall, we find that 88% of employees in the region experience multiple violations of their labour rights. When restricting to longitudinal data, we observe that the multidimensional labour violations index is persistent over time, with only 21.5% of employees changing status over the three-year period. These results highlight the need to foster both workers’ wellbeing and the rule of law in West Africa. About the authors

the multidimensional labour violations index is persistent over time, with only 21.5% of employees changing status over the three-year period. These results highlight the need to foster both workers’ wellbeing and the rule of law in West Africa. About the authors Thierno Malick Diallo is Lecturer at Gaston Berger University, Senegal; and has been involved in research collaborations with the Partnership for Economic Policy, Kenya (PEP) and the African Economic Research Consortium (AERC). His research centres around labour economics, gender, poverty, and rural economics. Lucas Ronconi is Professor of Economics at the University of Buenos Aires, Argentina. He is also a research fellow at the Argentine National Research Council, CONICET, a non-resident research fellow at the Institute for the Study of Labor (IZA), Germany, and at PEP , Kenya. His main research interest is labour market institutions in developing countries, with a focus on enforcement.02 ILO Working Paper 152 Abstract 01 About the authors 01 X Introduction 05 X 1 Methodology and Data 07 Multidimensional Labour Violation Index (MVI) 07 Household Survey 08 Legal Analysis 11 Benin 11 Burkina Faso 12 Côte d’Ivoire 12 Guinea-Bissau 12 Mali 13 Niger 13 Senegal 13 Togo 14 X 2 Measures of Compliance 15 Descriptive statistics 15 Econometric analysis 19 X 3 Persistence 22 X Conclusion 24 Appendix - EHCVM 25 References 26 Acknowledgements 28 Table of contents03 ILO Working Paper 152 List of Figures Figure 1 – Multiple Violations Index (MVI) weighting 07 Figure 2 – Multidimensional Labour Violations Index, whole sample 17 Figure 3 – Multidimensional Labour Violations Index, by country-year 18

Table of contents03 ILO Working Paper 152 List of Figures Figure 1 – Multiple Violations Index (MVI) weighting 07 Figure 2 – Multidimensional Labour Violations Index, whole sample 17 Figure 3 – Multidimensional Labour Violations Index, by country-year 18 Figure 4 – Persistence of Multiple Labour Violations Index (MVI) over time 2304 ILO Working Paper 152 List of Tables Table 1. Labour force descriptive statistics, WAEMU countries (2019 & 2022) 08 Table 2. Characteristics of Employees, WAEMU countries (2019 & 2022) 10 Table 3. Prevalence of labour law violations in West African countries (%) 16 Table 4 – Estimates of the determinants of Multidimensional Labour Violations Index 21 Table 5 – Persistence of MVI by sex 2305 ILO Working Paper 152 X Introduction Decent working conditions for all workers is a central objective of modern societies. To achieve this objective, governments implement labour regulations that provide protections to workers. Much of the academic and policy discussions have centred on how much labour protection is granted by the law, also called the level of labour protection. Some of the key contributions supporting the view that labour protection increases welfare are Galli and Kucera (2004), Deakin and Sarkar (2008), Deakin et al. (2014), and Adams et al. (2019); while others argue that the effects are mostly negative (Botero et al., 2004; Djankov and Ramalho, 2009; Heckman and Pages, 2004). However, how the law is observed in practice – the degree of compliance – has paradoxically received less attention. This is particularly problematic in low-and-middle-income countries (LMIC), where there is usually a large gap between statutory requirements and their effective implementation (Diallo and Ronconi, 2024). Noncompliance with labour protections undermines the credibility of governments and respect for the law. More importantly, it hurts the well-being of

where there is usually a large gap between statutory requirements and their effective implementation (Diallo and Ronconi, 2024). Noncompliance with labour protections undermines the credibility of governments and respect for the law. More importantly, it hurts the well-being of workers and their families. Noncompliance with legally mandated labour benefits is the focus of the paper. We use the terms ‘violation’ as synonymous with noncompliance. We compute a multidimensional index of labour violations that covers minimum wages, leave benefits and social security in eight West African countries (Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, Togo); and measures the persistence of violations over time. That is, we cover variation in compliance across space, time, and type of regulation. This allows for a much richer analysis of worker vulnerability. Because informality is a dominant feature of the labour market in these countries, the initial literature was not concerned with labour violations, but rather with the low productivity of the informal sector, which was understood as a manifestation of poverty and underdevelopment. For this reason, early definitions of informality (or rather, the informal sector) included subsistence self-employees (own-account workers) as well as employees working in small firms, regardless of access to labour benefits (Hart 1973; Tokman 1978). More recently, research on informality has begun estimating the extent of violation of workers’ rights, measured either as the share of employees without contributions to legally mandated social security, or as the share of employees with earnings below the minimum wage (Williams and Schneider (2016), Gindling et al. (2015), Rani et al. (2013), and Ronconi (2010)). Few of these studies, however, have covered African countries due to data restrictions; but this is changing thanks to the work of Bhorat et al. (2015, 2017) who cover Kenya, Namibia, Mali, South Africa, Tanzania, Uganda, and Zambia; and Badaoui and

tries due to data restrictions; but this is changing thanks to the work of Bhorat et al. (2015, 2017) who cover Kenya, Namibia, Mali, South Africa, Tanzania, Uganda, and Zambia; and Badaoui and Walsh (2022) who cover Burkina Faso, Côte d’Ivoire, Mali, and Niger. See also the estimates in ILO (2020) for the whole continent. Having precise estimates of compliance is a clear improvement in our understanding of informality. But a common characteristic of this literature is the focus on only one dimension at a time (typically minimum wages or social security). This is limited from a policy perspective since additional statutory protections, such as leave benefits, hours of work, safe and healthy working conditions, and protection against unfair dismissal, are of concern to workers. Once we take this broader perspective several questions emerge. Do labour violations co-occur? Are the same workers who suffer wage theft also excluded from social security and paid leave; that is, are workers subject to multiple violations of their rights? To answer these questions, it is necessary to analyse many dimensions and their distribution across workers.06 ILO Working Paper 152 This is the approach followed by Bhorat et al. (2021). They develop the ‘Multidimensional Labor Violation Index’ (MVI) by adapting the methods used in the multidimensional poverty literature, or MPI (Alkire and Foster, 2011; Alkire and Santos, 2013). The idea is simple; just as poverty has many dimensions beyond low income (such as education, food and housing), so does labour violation. Thus, the MVI is an index that covers several labour rights and benefits. Bhorat et al. (2021) apply the methodology to South Africa using data from the Quarterly Labour Force Surveys in 2014. The data allows compliance to be explored through minimum wages, leave benefits, written contracts, unemployment insurance and hours worked. The authors find substantial inequality in the distribution of violations. Our key contribution to this incipient literature is to add the time dimension. By following workers

in 2014. The data allows compliance to be explored through minimum wages, leave benefits, written contracts, unemployment insurance and hours worked. The authors find substantial inequality in the distribution of violations. Our key contribution to this incipient literature is to add the time dimension. By following workers over time, we can explore the persistence of labour violations. Do we observe many transitions, wherein workers’ labour rights are violated at t but then enjoy better working conditions at t+1? Or are violations more structural and persistent over time, wherein workers whose rights are violated upon entering the labour market remain in that status for the rest of their professional life? This is a key aspect of inequality that, to the best of our knowledge, has not been explored in the context of multidimensional labour violations. Empirically, in this paper we: (1) use microdata to compute individual-level measures of compliance covering several labour benefits (i.e., minimum wages, paid leave and social security), allowing the construction of a multidimensional index; (2) cover eight West African countries (Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, and Togo) at two points in time (2019 and 2022); (3) compute a matrix of transitions in labour status for those workers that are followed over time, measuring the persistence of violations; and (4) econometrically explore how compliance varies across workers’ and firms’ characteristics, sectors, and occupations. This paper attempts to contribute to the literature on compliance with labour protections. First, for many African countries, there is very little to no research measuring the extent of compliance with labour regulations beyond the minimum wage (such as social security, parental leave, or paid vacation). Second, some countries we cover – such as Benin, Guinea-Bissau, and Togo – are heavily under-researched in the fields of Labour and Development Economics (Das et al. 2013). Third, computing multidimensional indexes is conceptually important since it allows exploring whether nonregistered workers are, at least to some extent, compensated for lack of access to

heavily under-researched in the fields of Labour and Development Economics (Das et al. 2013). Third, computing multidimensional indexes is conceptually important since it allows exploring whether nonregistered workers are, at least to some extent, compensated for lack of access to social security by receiving other fringe benefits and/or higher wages. In terms of policy recommendations, the paper aims to identify specific sectors and occupations where compliance is particularly low, thereby guiding policymakers to target their efforts more effectively. By including the time component, the research explores structural issues; and by incorporating a comparative analysis across multiple West African countries, the research can also provide insights into regional patterns and best practices that can be adopted by policymakers to enhance labour standards. The paper is organized as follows: In section 2.1 we describe the construction of the multidimensional labour violation index; in sections 2.2 and 2.3 we describe the two inputs used to empirically operationalize the index: (i) household surveys and (ii) legal analysis. In section 3 we present the estimates of labour exclusion using the whole sample of stacked household surveys, as well as an econometric analysis of its determinants. The section also analyses persistence using longitudinal data.07 ILO Working Paper 152 X 1 Methodology and Data

Multidimensional Labour Violation Index (MVI) The MVI uses the worker as the unit of analysis. The construction of the index is simple. A worker is categorized as suffering a violation for a given legally mandated labour benefit if she/he fails to receive it. We do this categorization by comparing real working conditions (obtained from household surveys), with labour standards (set in the labour codes and laws). We explain the details in the next section. Then, we create a multidimensional Indexi variable, where the worker i is assigned a ‘violation score’, defined as the weighted sum of the number of violations of laws mandating labour benefits. Following the MPI literature, we assign equal weighting to each of the three dimensions we cover (i.e., wages, paid leave, and social security); and each labour inis assigned a ‘violation score’, defined as the weighted sum of the number of violations of laws mandating labour benefits. Following the MPI literature, we assign equal weighting to each of the three dimensions we cover (i.e., wages, paid leave, and social security); and each labour indicator within the dimension is also weighted equally. Figure 1 illustrates the weighting scheme, and equation (1) defines the formula. X Figure 1 – Multiple Violations Index (MVI) weighting For each worker in the sample i, the Index is a weighted sum of the five labour indicators (lj) where the weights (wj) are described in Figure 1. Then, and following the MPI literature, we categorize an employee as ‘multiply deprived’, if her/his score exceeds the 0.5 threshold. An important clarification note is necessary here. Settling the threshold is inherently normative. Choosing a value close to 0 implies categorizing as ‘multiply deprived’ workers who do not receive one, or more, legally mandated labour benefit. This threshold could be justified and motivated by noting that workers are usually more vulnerable than employers; and by emphasizing the concept of “rule of law” wherein violating the law is intrinsically wrong. In this paper, however, we ad hoc set the threshold at a much higher value, 0.5, implying that even some workers who are unlawfully denied benefits from two out of five regulations are categorized as not ‘multiply08 ILO Working Paper 152 deprived’. We follow this approach simply because the prevalence of violations of workers’ rights is so high in Western Africa that choosing a lower threshold implies having an almost empty set of not ‘multiply deprived’ workers, making the categorization almost useless. Finally, we compute the overall Violation Rate, which is defined as the share of employees that are multiply deprived. The next two subsections present the two key inputs we use to construct measures of compliance with labour standards: (1) the harmonised household survey conducted in 2019 and 2022; and (2) legal analysis of labour regulations in the eight African countries we cover. Household Survey

The next two subsections present the two key inputs we use to construct measures of compliance with labour standards: (1) the harmonised household survey conducted in 2019 and 2022; and (2) legal analysis of labour regulations in the eight African countries we cover. Household Survey The Harmonised Survey of Household Living Conditions (Enquête Harmonisée sur les Conditions de Vie des Ménages, EHCVM) was conducted in two rounds across the West Africa Economic Monetary Union (WAEMU) countries: Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, and Togo. The first round took place in 2018/19, and the second round in 2021/22. The data was collected by each national statistical agency, using a harmonised questionnaire. The collected data is a nationally representative sample of approximately 20,000 individuals per country and year. More details are available in the Appendix. Particularly crucial for our study is that the individual-level data includes an extensive labour module, which provides rich information about working conditions in the WAEMU region. Another advantage is that some of the sampled households and individuals are tracked over time. Our analysis is based on two different samples. First, we examine the legal coverage using a cross-sectional approach and focusing on household members over the age of 15 in each survey round. The analysis is in section 3. Second, we focus on the dynamic nature of labour compliance by constructing a panel data set that consists of household members above 15 years of age who work as employees across the panel rounds. For this purpose, we retain individuals for whom we have two observations over time to investigate the dynamics. This is shown below in section 4. Table 1 presents basic demographic characteristics for individuals aged 15 to 64 who participate in the labour force. The sample is obtained by appending all eight countries and years of the EHCVM survey. X Table 1. Labour force descriptive statistics, WAEMU countries (2019 & 2022) Variable Obs. Mean Std. Dev. Min Max

pate in the labour force. The sample is obtained by appending all eight countries and years of the EHCVM survey. X Table 1. Labour force descriptive statistics, WAEMU countries (2019 & 2022) Variable Obs. Mean Std. Dev. Min Max Age 231,334 35.9 12.79 15 64 Gender Male 231,334 .528 .499 0 1 Female 231,334 .472 .499 0 1 Education Primary education or less 199,516 .925 .264 0 1 Secondary education 199,516 .057 .232 0 1 Tertiary education 199,516 .018 .135 0 1 Location Urban areas 228,884 .386 .487 0 109 ILO Working Paper 152 Variable Obs. Mean Std. Dev. Min Max Rural areas 228,884 .614 .487 0 1 Occupational status Senior manager 104,701 .016 .124 0 1 Skilled workers 104,701 .151 .358 0 1 Unskilled workers 104,701 .833 .373 0 1 Employment sector Public sector 231,265 .038 .192 0 1 Private sector 231,265 .962 .192 0 1 Industry type Agriculture 231,334 .53 .499 0 1 Mining 231,334 .008 .089 0 1 Manufacturing 231,334 .113 .317 0 1 Services 231,334 .348 .476 0 1

Source: Own calculations using EHCVM 2018/19 and 2021/22. Notes: The table includes data for individuals who participate in the labour force in the eight WAEMU countries (Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, and Togo),

in the two years when the EHCVM survey was conducted (2019 and 2022). The average age, and particularly the average level of human capital (education, skills and experience), is low compared to the labour force in other countries.1 Also, the share of the workforce residing in urban areas, and the size of the public sector, are both relatively small. Measuring education is complex in the region because many Muslim individuals attend a madrassa institution instead of a formal school. While children attending a madrassa institution may learn several other topics beyond the religion of Islam, they may report “no primary education” in the household survey, or do not answer the question, since they did not attend a formal school. In Table 2 we restrict the sample to employees and present their basic demographic characteristics. The first important aspect to note is that employees represent a small share of the labour force in all analysed countries, starting from as low as 10% in Niger. Even in Senegal, the country in the region where employees reach the maximum level of 30% of the labour force, it is still much lower than in other parts of the world. This is because most workers in West Africa are engaged in subsistence agriculture or precarious self-employment. 1 See World Development Indicators https://databank.worldbank.org/source/world-development-indicators and ILOSTAT https://ilostat. ilo.org/data/.10 ILO Working Paper 152 X Table 2. Characteristics of Employees, WAEMU countries (2019 & 2022) Sample Female Secondary educ. or more† Share employee No. Employees surveyedMean SD Mean SD Benin 2019 0.381 0.486 0.380 0.486 0.164 3,175

2022 0.410 0.492 0.320 0.466 0.166 2,800 Burkina Faso 2019 0.355 0.479 0.239 0.423 0.138 3,303 2022 0.366 0.482 0.200 0.401 0.262 1,799 Côte d'Ivoire 2019 0.296 0.457 0.236 0.425 0.231 5,853 2022 0.318 0.466 0.229 0.420 0.181 3,413 Guinea-Bissau 2019 0.336 0.472 0.380 0.485 0.209 4,164 2022 0.331 0.471 0.508 0.500 0.179 3,033 Mali 2019 0.269 0.443 0.323 0.468 0.220 3,239 2022 0.248 0.432 0.288 0.453 0.236 2,433 Niger 2019 0.222 0.416 0.304 0.460 0.105 1,624 2022 0.233 0.423 0.279 0.448 0.099 1,398 Senegal 2019 0.280 0.449 0.168 0.374 0.338 6,959 2022 0.278 0.448 0.194 0.395 0.261 4,574 Togo 2019 0.316 0.465 0.498 0.500 0.144 1,925 2022 0.386 0.487 0.551 0.498 0.136 1,274

WAEMU Total 2019 0.308 0.282 0.450 0.499 0.198 30,242 2022 0.318 0.300 0.458 0.471 0.187 20,724

Source: Own calculations using EHCVM 2018/19 and 2021/22. Notes: †The education variable is only available for 33,099 employees in the whole sample.

Women are less represented in salaried employment. While they represent almost half (47%) of the whole labour force, the share of employees that are female is less than one third. Education, as discussed above, is not easy to measure because of the existence of the madrassa institutions. This is a reason why Senegal, a country with a relatively large Muslim population, appears to have very low levels of formal education attainment (i.e., less than 20% of employees with secondary education or more). Employees, compared to other members of the labour force, are substantially more educated (i.e., the share with secondary education or more is below 10% in the whole labour force as shown in Table 1, compared to almost 45% among employees as shown in Table 2).11 ILO Working Paper 152 Legal Analysis This section determines the legal coverage, as well as the level, of each of the labour regulations available in the household survey. We first present some general clarifications and then analyse each labour regulation in each country. Since the EHCVM was carried out in 2018/2019 and 2021/22, the legal analysis focuses on the legal coverage applicable during both time frames. ● For minimum wages (MW), we focus on the statutory minimum wage, that is, the minimum floor set by the government. We do not cover the wages set via collective bargaining. For employees who report working fewer hours that the regular workweek, we assume they are legally entitled to a MW that is proportional. For example, if the MW is $100 per week, the regular workweek is 40 hours, and the employee reports working 30 hours per week, then,

employees who report working fewer hours that the regular workweek, we assume they are legally entitled to a MW that is proportional. For example, if the MW is $100 per week, the regular workweek is 40 hours, and the employee reports working 30 hours per week, then, we assume that the MW for that worker is $75 per week. ● In most countries, the labour code provides for paid annual leave for all workers after one year of actual service. Therefore, our analysis of compliance (with respect to this benefit), only includes employees with one year or more of tenure.2 ● We exclude severance pay from the analysis –despite being compulsory for most employees in the region –since the information in the household survey is insufficient to precisely determine coverage. In all cases, the main source of information we use is each country’s legislation. We also analysed several ILO documents and databases to determine legal coverage including EPLex and ISSA; and Wage Indicator data.3 In case of doubts, we consulted with legal scholars. Benin ● Statutory minimum wage is determined as a monthly wage. It is based on the legal working hours, which are 46 hours per week for agricultural workers and 40 hours per week for all other workers. From 2014 to 2022 the MW was 40,000 CFA West African Franc per month. ● Annual Paid Leave: The labour code (article 158) provides for paid annual leave for all workers after one year of actual service. ● Maternity/Paternity Paid Leave: Employed women are entitled to 14 weeks of maternity leave with full pay. There is no specific law mandating paternity leave. ● Paid Sick Leave: The labour code provides for paid sick leave for all workers (duration and amount of payment depends on the length of service). ● Contributions to Social Security: All employees are entitled except agricultural workers.4 2 It is not obvious whether to include or not employees with less than one year of tenure. On the one hand, an employee with less than one year of tenure who reports receiving “no annual paid leave” in the survey, should be categorized as a case of non-compliance. This

2 It is not obvious whether to include or not employees with less than one year of tenure. On the one hand, an employee with less than one year of tenure who reports receiving “no annual paid leave” in the survey, should be categorized as a case of non-compliance. This is because every worker –regardless of tenure– is entitled to (proportional) annual paid leave. For example, if a worker is dismissed before completing one year of service, he or she is entitled to receive an amount of money proportional to the time worked during the year. On the other hand, it is quite li