Increasing the employment of women through flexible work arrangements Statements and Comments

Transcription

1 Increasing the employment of women through flexible work arrangements Statements and Comments 1. Background İnsan TUNALI Koç University In 2004 the countrywide employment rate (15+) in Turkey was 22.9 percent for females, and 64.7 percent for males. The figures in rural areas were considerably higher (F: 35.5, M: 69.3) than in urban areas (F: 15, M: 62). The differences by location reflect the vital role of agriculture in employment (estimated at around 7.4 million, amounting to 34 percent of the total) and the continuing importance of family farming practices. The differences by gender reflect the predominant form of division of labour in the household, which assigns the male head the breadwinning role, and his wife the care taking role. The public sector employed 13.6 percent of the work force (21.8 million) in Females accounted for 20.9 percent of public sector employment, 17.2 percent of non-agricultural, and 44.6 percent of agricultural employment. In 2004 the unemployment rate was 10.3 percent countrywide (F: 9.7, M: 10.5), and significantly higher in urban areas (F: 17.9, M: 12.5) compared to rural areas (F: 3.2, M: 7.3), especially in the case of women. The high unemployment/low employment rates recorded in urban areas reflect the lingering effects of the economic crises in 2000 and 2001, which had a huge toll in the labour market (see Tunalı et al., 2003). Although the economy rebounded in 2002 and grew at an average annual rate of 7.5 percent over the period, the recovery has not produced job growth. There is evidence that the length of the work week increased during this time, and reached a record 50.7 hours in 2003 in the case of regular wage and salary earners (F: 47, M: 51.7). A marked increase in informal employment arrangements has been another consequence. In 2004, 20.9 percent of all regular wage and salary workers, and 34 percent of the non-agricultural work force were outside the coverage of the social security system. The female LFPR in Turkey has been declining since the 1950s. This period has been characterized by massive rural-urban migration, and an increased market orientation as captured by the fraction of the paid workforce in total employment. Nonetheless in 2004 nearly half of all employed females were classified as unpaid family workers, while 11 percent were self-employed, and only 39 percent were wage and salary workers. The share of the latter group in the female work force in urban areas (roughly 2.3 million) was 79 percent. Although the secular decline in female participation and the shift to market work fits the historical patterns observed elsewhere (see Mammen and Paxson, 2000), why the female LFPR in urban areas has remained under 20 percent for the last 15 years, despite favourable trends in education, fertility, and increased use of labour saving appliances at home, is harder to explain. Tunalı et al. (2003, pp ) offer potential explanations, and point out the absence of flexible employment arrangements as a detrimental factor. Consequently the Norwegian case study and the debate it is likely to engender, provide a good opportunity for rethinking an important dimension of the female labour supply puzzle in Turkey. 1

2 2. The policy context Implicit government policy regarding fertility since the 1960s may be described as family planning/fertility control oriented (Shorter, 1995). Significant reductions in fertility have been observed in the last decade or so (TFR was 4.3 in 1978, 2.65 in 1993, 2.61 in 1998, 2.23 in 2003). There are no government administered cash benefits tied to family size, nor public child care facilities. The share of part time workers is extremely low in Turkey. According to the latest data (from HLFS 2003), around 4 percent of male, and 9 percent of regular female wage and salary earners worked 35 hours or less during the reference week. As shown in Figure 1, more than half of those in the sample reported hours in the range, and the remainder reported working 50 hours or more. 1 This pattern is attributable to the fact that social security legislation was written with the full time male bread winner in mind. Simply put, it was not possible to link tax and benefit payments directly with hours of work, and consequently firms were less than eager to hire part time workers. The new Labour Law (no enacted in 2003) was designed, in part, to address employer concerns about lack of flexibility in the nature of the employment relationship, and intensity of work during the day/week/month. It introduces many atypical employment arrangements inspired by models used elsewhere. 2 In addition to formalizing part time and short term employment, the new law also extends maternity leave privileges, and includes a special clause for provision of nursing and day care facilities. The proponents of the new law argue that it will make it possible for women to balance work and family responsibilities. Brief summaries of the clauses that pertain to the policy measures (which comprise the subject matter of the peer review) are given below. Paid leave arrangements: Paid maternity leave lasts for a total of 16 weeks, and is normally divided evenly (8 weeks each) before and after delivery. If health of the female worker permits, the leave can be postponed for up to three weeks before term, so that up to eleven weeks can be used after birth. The leave can be extended for up to six months without pay, without risk of loss of job. Fathers are not eligible for leave. Working time: Length of the normal work week remains set at 45 hours per week. However, the new law allows substantial fluctuations without obliging the employer to pay for overtime, providing the parties agree to it, and the maximum does not exceed 11 hours per day. In fact the work week can be as long as 55 (66) hours per week in an establishment that work five (six) days of the week, however the average over a two month period cannot exceed 45 hours. Part time work: Work which lasts no longer than two-thirds of the hours set in the reference job is considered part time work. Since the reference job will typically have normal hours, the 1 These calculations are based on the data reported on the web page of the State Institute of Statistics. The data are reported in ranges shown in Figure 1. According to the common definition adopted by the OECD, the share of part time employment (30 hours or less) among dependent employment in Turkey in 2003 was 1.8 percent for males, and 5.4 percent for females. 2 The newly introduced categories and terminology include the following [Turkish originals are in brackets]: part time work [kısmi süreli çalışma], work upon recall (by employer) [çağrı üzerine çalışma], temporary (on loan) employment relationship [geçici (ödünç) iş ilişkisi], compensatory work [telafi çalışması], fixed term work [belirli süreli çalışma]. Descriptions are provided in the Appendix. 2

3 cut-off for part time work will typically be 30 hours per week. The circular issued by the Ministry of Labour and Social Security (dated ) indicates that each 7.5 hour portion of part time work will count as one full day of work for the purposes of calculating the social security premium (section V, paragraph no. 4). Nurseries, day care facilities: Article no. 88 indicates that regulations that govern the limitations to be placed on the conditions of work for pregnant and nursing women, and the provision of nurseries and day care facilities will be issued by the Ministry of Labour and Social Security upon consultation with the Ministry of Health. According to the current regulation, establishments that employ female workers (regardless of age and marital status) have to provide a nursing room to facilitate nursing of (0-1 year old) infants. In establishments that employ more than 150 female workers, a day care facility that addresses the needs of preschool (ages 0-6 years old) children, has to be provided instead (section 6, article 15). Although the new law marks a departure from a paternalistic philosophy to an egalitarian one (e.g. the ban on employment of women in night shifts of manufacturing establishments has been lifted), some of the stipulations (such as those in Article 88) may backfire and result in discriminatory practices against female workers unless employment practices of firms can be effectively monitored. We end this section with a translation of the equal treatment article. Equal treatment principle (Article 5): There can be no discrimination in the employment relationship based on language, race, sex, political views, philosophy of beliefs, religion, sect and similar reasons. Unless there are fundamental reasons, employer cannot treat a part time worker differently from a full time worker, nor a worker on an indefinite contract differently from a worker on a fixed term contract. Unless there are reasons due to biology or nature of work, employer cannot treat a worker differently in the manner the contract is written, applied, or terminated, directly or indirectly because of pregnancy or sex. For same or equivalent work, a lower wage cannot be set because of sex. The provision of special protective clauses because of sex cannot be grounds for a lower wage. 3. Potential for transferability In 2003, the female LFPR (15+) in urban areas for single women was more than twice that for married women (31.1 vs percent). Single women accounted for 23.8 percent of the (15+) population, but 41.4 percent of the labour force. By contrast married women accounted for 65.8 percent of the population, and 51.2 percent of the labour force. Divorced women had the highest LFPR at 42 percent, and comprised 5.1 percent of the population. Thus entry into marriage appears to usher in withdrawal from the labour force, and the likelihood of return at a later time appears to be very low unless the union ends in a divorce. 3 Taken at face value, these stylized facts suggest that adoption of family friendly policies might have a positive impact on female labour supply, providing they address the constraints that married women face. Notably there is supportive micro econometric evidence which associates lower 3 Since longitudinal data are not available, in drawing this tentative inference we rely on cross-section data. 3

4 participation probabilities with the presence of young children, and higher participation probabilities with the availability of help from extended family members (Tunalı and Başlevent, 2005) Likely impact of the leave and child care policy: By way of assessing the role of government policies, consider Table 3 of the policy paper submitted by the two Norwegian Ministries. According to the country data compiled there, total family support amounted to 1.1 percent of the GDP in Turkey in 1999, a figure which is significantly below the 2001 averages reported for the OECD (1.8 percent) and the EU-15 (2.2 percent). Based on this, one might argue that lack of sufficient support could be responsible for the withdrawal of married females from the labour force. Note that in the row for Turkey, support for maternity leave and day care services are given as 0 percent. Yet the old laws did have maternal leave clauses (that allowed 9 weeks of paid leave for civil servants and 12 weeks for others) and some collective bargaining arrangements stipulated day care provisions in large establishments. Thus it would be fair to say that some support was available and the new law extends the available support. This being the case, the impact of the extended leave and child care policy might be marginal Likely impact of flexible working time arrangements: In Turkey the ubiquity of informal employment arrangements has often been pointed out as a potential source of flexibility. In this vein, one form of employment that the Household Labour Force Survey keeps track of is casual work, which is defined as daily, seasonal, or temporary work. Remarkably in 2003, share of casual employees who worked 35 hours or less per week was 38.4 percent in the case of females, and 13 percent in the case of males (see Figure 2). At that time casual employees accounted for less than 6 percent of the female, and less than 9 percent of the male work force. Although these data suggest that there is demand for flexible employment arrangements, it is hard to gauge the extent to which they will be accepted by the broader segment of the labour force. In fact a second piece of evidence from the Household Labour Force Survey, based on a question on work intensity preferences of job seekers, suggests that the demand for part time arrangements is quite limited. In 2000 (before the economic crisis had its impact), only 5 percent of the females, and 3 percent of the males indicated that they were looking for part time jobs. Individuals searching for full term jobs accounted for 82 percent of the respondents from either sex. The remainder said they were indifferent between part or full time jobs. In 2003, the share of part time job seekers was respectively 3 and 1 percent among female and male respondents, while the share of full time job seekers remained around percent. It should be underscored that these statistics were collected during a time when legislation was not favourable for part time employment. The new law offers a balanced stance. Since the rules and regulations needed for full implementation were not completed until mid 2004, it will take some time before the impact of the new law can be assessed. 4

5 4. Reform of the Social Security System The age threshold for full retirement in Turkey has been implicated as being one of the most generous in the world. Although the retirement age was increased in 1999 (to 58 for females and 60 for males) this did not alleviate the threat to the solvency of the social security system. The proposed legislation (which is about to reach the floor of the Parliament) is designed to reduce inefficiencies (by combining the three existing branches under one administration) and actuarial concerns (by adjusting premium payments and pension benefits). The proposal also establishes the contours of a new universal health insurance and social welfare system. By altering the existing system of incentives, the new legislation will surely have a major impact on labour market outcomes. It is impossible to do justice to the subject in the confines of this paper. However, suffice it to say that passage of the new social security law will be an important step in aligning legislation in Turkey with that in the EU. References Mammen, K. and C. Paxon (2000) Women s Work and Economic Development, Journal of Economic Perspectives, 14 (4): Shorter, F. (1995) The Crisis of Population Knowledge in Turkey, New Perspectives on Turkey, no.12 (Spring): TİSK (2004) Employment, Female Labour Force, and the New Labour Law Symposium (held in Muğla, 5 December 2003; in Turkish). Ankara: Turkish Confederation of Employers Associations. Tunalı, İ., with H. Ercan, C. Başlevent, O.D. Öztürk and U. Akçiğit (2004) Background Study on Labour Market and Employment in Turkey. European Training Foundation, Torino, Italy. Tunalı, İ., and C. Başlevent (2005) Female Labour Supply in Turkey. Mimeo (April) 45pp. 5

6 Figure 1. Hours of work during the reference week, regular Wage & Salary workers 0,40 Regular Wage & Salary Workers 0,35 0,30 0,25 0,20 M F 0,15 0,10 0,05 0, Source: Household Labour Force Survey, Figure 2. Hours of work during the reference week, Casual workers 0,40 Casual Workers 0,35 0,30 0,25 0,20 0,15 M F 0,10 0,05 0, Source: Household Labour Force Survey,

7 Appendix. Newly introduced categories and terminology in Work Law no Part time work [kısmi süreli çalışma]: Allows weekly hours of work to be flexible. Part time work is defined as work which lasts for no more than two-thirds of the weekly hours stipulated for the full time worker who serves as the reference. For most jobs, full time work is 45 hours; thus practically speaking, part time work amounts to 30 hours or less per week. Statement of reason of article no. 14 indicates that objectives stipulated in EU directive 97/81 have been adopted. Work on recall (by employer) [çağrı üzerine çalışma]: Applicable in situations where worker does not have to be present in the workplace at all times. Allows employer to recall worker (with whom a written contract has been signed) when the need for worker arises. In the absence of a mutually agreed contractual stipulation, employer has the obligation to compensate the worker for 20 hours of pay per week. The inspiration for this article is the German Employment Promotion Law. Temporary (on loan) work relationship [geçici (ödünç) iş ilişkisi]: Employer can loan a worker to another employer upon written acceptance by worker. The temporary work relation is for a fixed term stipulated in a written contract, and can only be renewed once. Compensatory work [telafi çalışması]: If the normal pace of work is interrupted for reasons beyond the employer s control, because of extended holidays, or because of a leave of absence initiated by the worker, employer can require compensatory work without pay within two months of the interruption. Fixed term work [belirli süreli çalışma]: Applies to situations in which work lasts for a fixed term because of its nature or intended objectives. Fixed term contracts cannot be rewritten unless there is a sound reason. Repeated (chain) contracts that have a sound base remain as fixed term contracts under the law. 7