Labor Market Structure in Indonesia: Segregation or Integration of Occupations?

This research discusses occupational segregation of workers based on gender in 2015 2018 in Indonesia. This study also aims to measure the level of dissimilarity or how much segregation occurs in the labor market in Indonesia. The sample data used were both males and females labor includes residents aged 15 years and over, namely workers aged 35-44 years. Measurement of occupational segregation is based on between occupational groups and provinces which is explained by a dissimilarity index. A dissimilarity index was constructed to measure occupational segregation. First, the percentage of all workers in each province where each occupation group is calculated. This index is then half the absolute value of the difference between the specific locations of the distribution. The occupational structure in Indonesia during the observed period shows that there is no occupational segregation based on gender where the index number D tends to be closer to zero. The occupation of the leadership and management of men and women workers is the most integrated. From the index number D, it means that female workers are less segregated in all types of occupation in the labor market in Indonesia. Integration is showing signs of increasing in Professional occupations, technicians and the like; Administration; Sales Business; and agricultural, forestry, hunting and fishery businesses. Y. Yunisvita, Muhammad Teguh, Rosmiyati Chodijah, I. Imelda, Arika Kurniawan Faculty of Economics, Universitas Sriwijaya How to Cite: Yunisvita, Y., Teguh, M., Chodijah, R., Imelda, I., & Kurniawan, A. (2021). Labor Market Structure in Indonesia: Segregation or Integration of Occupations?. Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 22(1), 42-53. doi:https://doi.org/10.23917/jep.v22i1.12791 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v22i1.12791 Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 22 (1), 2021, 42-53 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 43 the increasing participation of females in the labor market is accompanied by a decrease in occupational segregation based on gender. Occupational gender segregation is complex, and increasing women’s labor market participation is not a straightforward path to creating an occupational structure that is less segregated by gender (Gedikli, 2020). About half of women would need to shift into a new occupation to eliminate all occupational segregation by gender (Weeden, Newhart, and Gelbgiser, 2018). Women combine continuing labor market participation with home and family responsibilities particularly during the childcare years. It can be argued that, in the context of women’s work-family preferences, inferior conditions, notably lower pay, are not necessarily evidence of discrimination or disadvantage (Connolly and Gregory, 2008). The level of participation in Indonesia shows a fluctuating situation but the trend is increasing (Table 1). In 2018 males LFPR was 82.68 percent, increasing slightly to 83.13 in the following year. Meanwhile, females LFPR in the same period did not change at 51.88 percent. In general, the LFPR (Labor Force Participation Rate) for males is still much higher than females LFPR, around 1.5 times females LFPR. Based on occupation, female workers dominate in types of work (1) professional, technicians and related occupations; (3) Clerical and related occupations; (4) Sales worker and (5) Services worker. Male workers are more dominant in the type of work (2) Managerial and supervisory occupations; (6) Agriculture, forestry, hunting and fishing workers and laborers; (7) Production workers, operation of machinery workers; and (8) others. The dominance of female in this type of work is 2 times more than that of male workers in this type of work (Table 2). The highest percentage of female workers is in the type of work (4) Clerical and related occupations, namely 26.53 percent, while male workers are mostly in the type of work (7) Production workers, operation of machinery workers namely 37.37 percent. Several empirical studies provide evidence that females prefer working in certain jobs and firms, because they are associated with lower investment in job-specific training (Becker, 1971), less competitive environment (Niederle & Vesterlund, 2007), depreciation rate of human capital. lower levels (Görlich & de Grip, 2009, and Polachek, 1981), and more pleasant and familyfriendly working conditions (Bender, Donohue, and Heywood 2005; Budig and England, 2001). For this desire, the non-cash job characteristics of female seem willing to accept lower wages. Another explanation regarding occupational segregation is that the sex division is based on the different roles played by males and females due to biological factors. The allocation of men/women into already masculinised/feminised occupations may help in understanding this situation. In the case of women, they are basically concentrated in services-related occupations (more than 80% of female employment). The service sector is usually considered as offering more flexibility in the working day, allowing for greater compatibility of family responsibilities with paid work. It also allows for more interpersonal relationships, and possibly requires less physical effort. If working women choose occupations with certain preferred job attributes, then they may eventually reach a higher level of job satisfaction (Garcia, Mainar et al, 2015). Weeden, Newhart, and Gelbgiser (2018) stated occupational segregation factors include discrimination against women or mothers, gender-specific socialization, gender-linked traits or “natural” abilities, cultural beliefs about men and women’s competence and double standards of evaluation, the household division of labor, and workplace experiences (e.g., sexual harassment). Based on this, occupational segregation decreases in industries where the physical attributes of workers become less important. Empirical studies regarding the size and determinants of segregation measurement are mostly varied, where the occupational segregation level decreases, remains or increases (Fortin & Huberman, 2002). When the key determinants of occupational choice and therefore occupational segregation are education level, age, location, they have generally been shown to be significant (King, 1992; Sprigss and Williams, 1996). Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v22i1.12791 Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 22 (1), 2021, 42-53 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 44 Table 1. LFPR by Gender in Indonesia Gender August 2018 February 2019 August 2019 February 2020 August 2020 Male 82.68 83.18 83.13 83.82 82.41 Female 51.88 55.50 51.89 54.56 53.13 National 67.26 69.32 67.49 69.17 67.77 Source: BPS Indonesia, 2019 and 2020 Table 2 Percentage of Workers by Occupation and Gender in Indonesia Occupation KBJI Code Male Female Professional, technician and related occupations 1 5.77 10.59 Managerial and supervisory occupations 2 1.95 0.86 Clerical and related occupations 3 5.21 7,18 Sales worker 4 14.53 26.53 Services worker 5 4.22 9.66 Agriculture, forestry, hunting and fishing workers and laborers 6 27.96 24.35 Production workers, operation of machinery workers 7 37.57 20.68

the increasing participation of females in the labor market is accompanied by a decrease in occupational segregation based on gender. Occupational gender segregation is complex, and increasing women's labor market participation is not a straightforward path to creating an occupational structure that is less segregated by gender (Gedikli, 2020). About half of women would need to shift into a new occupation to eliminate all occupational segregation by gender (Weeden, Newhart, and Gelbgiser, 2018). Women combine continuing labor market participation with home and family responsibilities particularly during the childcare years. It can be argued that, in the context of women's work-family preferences, inferior conditions, notably lower pay, are not necessarily evidence of discrimination or disadvantage (Connolly and Gregory, 2008).
The level of participation in Indonesia shows a fluctuating situation but the trend is increasing (Table 1). In 2018 males LFPR was 82.68 percent, increasing slightly to 83.13 in the following year. Meanwhile, females LFPR in the same period did not change at 51.88 percent. In general, the LFPR (Labor Force Participation Rate) for males is still much higher than females LFPR, around 1.5 times females LFPR.
Based on occupation, female workers dominate in types of work (1) professional, technicians and related occupations; (3) Clerical and related occupations; (4) Sales worker and (5) Services worker. Male workers are more dominant in the type of work (2) Managerial and supervisory occupations; (6) Agriculture, forestry, hunting and fishing workers and laborers; (7) Production workers, operation of machinery workers; and (8) others.
The dominance of female in this type of work is 2 times more than that of male workers in this type of work ( Table 2). The highest percentage of female workers is in the type of work (4) Clerical and related occupations, namely 26.53 percent, while male workers are mostly in the type of work (7) Production workers, operation of machinery workers namely 37.37 percent.
Several empirical studies provide evidence that females prefer working in certain jobs and firms, because they are associated with lower investment in job-specific training (Becker, 1971), less competitive environment (Niederle & Vesterlund, 2007), depreciation rate of human capital. lower levels (Görlich &de Grip, 2009, andPolachek, 1981), and more pleasant and familyfriendly working conditions (Bender, Donohue, and Heywood 2005;Budig and England, 2001). For this desire, the non-cash job characteristics of female seem willing to accept lower wages.
Another explanation regarding occupational segregation is that the sex division is based on the different roles played by males and females due to biological factors. The allocation of men/women into already masculinised/feminised occupations may help in understanding this situation. In the case of women, they are basically concentrated in services-related occupations (more than 80% of female employment). The service sector is usually considered as offering more flexibility in the working day, allowing for greater compatibility of family responsibilities with paid work. It also allows for more interpersonal relationships, and possibly requires less physical effort. If working women choose occupations with certain preferred job attributes, then they may eventually reach a higher level of job satisfaction (Garcia, Mainar et al, 2015). Weeden, Newhart, and Gelbgiser (2018) stated occupational segregation factors include discrimination against women or mothers, gender-specific socialization, gender-linked traits or "natural" abilities, cultural beliefs about men and women's competence and double standards of evaluation, the household division of labor, and workplace experiences (e.g., sexual harassment). Based on this, occupational segregation decreases in industries where the physical attributes of workers become less important.
Empirical studies regarding the size and determinants of segregation measurement are mostly varied, where the occupational segregation level decreases, remains or increases (Fortin & Huberman, 2002). When the key determinants of occupational choice and therefore occupational segregation are education level, age, location, they have generally been shown to be significant (King, 1992;Sprigss and Williams, 1996). Well-known econometric studies on the role and contribution of gender shares in occupation to income include (Bayard et al., 2003), (Boraas & Rodgers III, 2003), (Johnson and Solon, 1986), (Kilbourne et al., 1994), (Macpherson and Hirsch, 1995), and Sorensen (1990) for the United States; (Busch & Holst, 2012) for Germany; Campos Soria and Ropero (2011) for Spain;(EC, 2002) for a combined sample of EU-12 countries; (Jurajda, 2003) for the Czech Republic and Slovakia; (Karamessini & Ioakimoglou, 2007) for Greece; and (Ogloblin, 1999) for Russia. These studies confirm lower earnings for females and males in femaledominated occupations. While the majority of studies show that the negative effects of working in female-dominated occupations on income are much more pronounced for female, studies by (Johnson and Solon, 1986) and (Sorensen, 1990) for the United States and (Busch & Holst, 2012) for Germany found that the negative effects were stronger for males. According to (Macpherson and Hirsch, 1995) study for the United States, the effect of gender composition on income is equally significant for female and male. The gender pay gap explained by occupational segregation varies widely across studies and countries. The share of gender in the occupational variable explains between 7 percent and 32 percent of the salary gap observed in studies in the United States, about 25 percent in cross-border studies of EU-12 countries (EC, 2002), also 50-56 percent in Greece (Karamessini & Ioakimoglou, 2007), and only 3 percent in Germany (Busch & Holst, 2012) Various explanations have been offered in the literature for the prevalence of occupational segregation. Key factors include: choice of field of study that results in segregation in education and ultimately in occupation; female's preference for the job allows for flexible or intermittent working hours; immeasurable worker differences; skills or job characteristics that may be gender correlated; stereotypes of employment as appropriate for females due to discriminatory recruitment practices or social norms; and hidden barriers to entry and bias in organizational practice. (Karamessini and Ioakimoglou, 2007) emphasize that the extent to which gender-based occupational segregation occurs depends on institutions, culture, and history. They argue that gender relations are influenced by wage fixing and collective bargaining practices. In the same vein, Ogloblin (1999) and Jurajda (2003) conclude that occupational segregation by gender in the labor markets of transition countries is more likely a legacy of communism in which central planners stereotype females into specific occupations and consequently shape social attitudes. Pailhé (2000) also argues that stereotypes are widely accepted as part of life by Central European women.
Meanwhile, the level of occupational segregation based on gender has shown a decreasing trend in most developed countries over the last few decades (Anker, 1998;Costa, 2000), the increase in female's employment in Spain has been accompanied by a steady increase in gender segregation so that Spain is currently a country where the gender differences in the distribution of employment among occupations are extraordinary (European Commission, 2009;García-Mainar et al., 2015) Occupational segregation is more severe in Latin America and the Caribbean but less so in South Asia and Sub-Saharan Africa. In Latin America the Duncan index is the highest (.53), followed by the Middle East and North Africa (.50), Europe and Central Asia (0.46), East Asia and the Pacific (.39), Sub-Saharan Africa (.33), and South Asia (.30). Low employment segregation in South Asia and Sub-Saharan Africa may be related to the limited granularity of employment codes in the agricultural sector (Das and Kotikula, 2019).
Occupational segregation is the result of "push" and "pull" factors that are rooted in social interactions and social structures. These factors include discrimination against women or mothers, gender-specific socialization, gender-related traits or "natural" abilities, cultural beliefs about male and female competence and multiple evaluation standards, household divisions of the workforce, experience in the workplace, government policies prohibiting discrimination in employment but allowing different wages for comparable work and worker family policies (Weeden, Kim. A; M. Newhart and D. Gelbgiser, 2018).
This research examines whether there is segregation or integration in the labor market in Indonesia and measures the degree of segregation or integration. This is important because the phenomenon of inequality still occurs, especially in many developing countries, including Indonesia.

Research Method
This study measures the level of dissimilarity index or how much occupational segregation that occurs in the labor market in Indonesia.The data presented was obtained from the National Labor Force Survey (Sakernas) which was carried out throughout the Republic of Indonesia in 2015-2018. The target sample size for Sakernas is 200,000 households, and is intended to produce estimates down to the district / municipality level. This publication uses a weighting of the results of population projections for 2010-2035. The sample data used were both male and female labor includes residents aged 15 years and over. For this study, the sample is reduced to workers who are in the "prime age" group, namely workers aged 35-44 years because they constitute a significant portion of the labor market (Gabriel & Schmitz, 2007).Types of work are divided into 8 groups according to KBJI (Classification of Indonesian Occupation Standards) 2002 and simplification is carried out in this study, namely: code 1 is Professional, technician and related occupations; code 2 Managerial and supervisory occupations; code 3 is Clerical and related occupations; code 4 Sales worker; code 5 is Services worker; code 6 is Agriculture, forestry, hunting and fishing workers and laborers; code 7 is Production workers, operation of machinery workers; Code 8 is others. Measurement of occupational segregation based on between occupational groups and provinces is described by (Duncan & Duncan, 1955). A dissimilarity index was constructed to measure occupational segregation. First, the percentage of all workers in each province where each occupation group is calculated. This index is then half the absolute value of the difference between the specific locations of the distribution, where ID is the Dissimilarity index; Fi is the number of female workers in occupation i, Mi is the number of male workers in occupation i; Fj is the number of female workers in j province; Mj is the number of male workers in the province j. The absolute value of the sum of the difference between the percentage distribution of males and females in each occupation is halved (because there are two groups of males and females) to produce values that range in the index from 0 (perfect integration) to 1 (perfect segregation).

Results
In Figure 1, it is shown that the types of occupation with codes 4 to 8 experience changes in the average wage of male workers with a negative slope. The largest decrease occurred in the occupation of Agriculture, forestry, hunting and fishing workers and laborers (code 6). During the 2015-2018 period, the same decline occurred in the occupation of Production workers, operation of machinery workers (code 7). Male workers in these two occupations experienced a decrease in wages of about half of the wages in 2015 which lasted until 2018. Meanwhile, the downward trend occurred in Services worker (code 5); Sales worker (code 4) and others (code 8) is 30 percent. Clerical and related occupations is a type of work that provides a constant average wage during that period. Meanwhile, Managerial and supervisory occupations (code 2) and Professional, technician and related occupations (code 1) showed a generally positive trend.
The pattern of changes in the wages of male workers based on occupation during this period also occurred for female workers. However, nominally female workers only receive wages between one-third and two-thirds of the male workers' wages ( Figure 2). To further analyze the segregation problem based on gender, it can be seen using the Dissimilarity Index (D-Index). Table 3 shows the results of the D index calculation based on gender. In the 2015-2018 period, the level of segregation decreased in professional and clerical occupations. In contrast to managerial and other occupations. Meanwhile, the level of segregation fluctuates in sales, services, agriculture and production occupations. The highest segregation is found in the occupation of production workers.
The occupational structure in Indonesia during the observed period shows that there is no occupational segregation based on gender where the index number D tends to be closer to zero. The occupation of the leadership and management of male and female workers is the most integrated. However, a trend segregation in this occupation from 2015 to 2018 increased as a whole. The same thing also happened to the service business occupation.
Generally on the index number of D means female workers are less segregated in all kinds of occupations in the labor market in Indonesia. Integration is showing signs of increasing in Professional, technician and related occupations; Clerical and related occupations; Sales worker; and Agriculture, forestry, hunting and fishing workers and laborers. In Professional occupation, if the annual average rate decreases since 2015 continue, it will take 15 years to achieve full integration, it will take 10 years for Clerical and related occupations; Meanwhile, for the occupation of the Agriculture, forestry, hunting and fishing workers and laborers, it will take 130 years.  Based on Table 4, Bengkulu, North Sulawesi and Gorontalo are the three provinces with the highest level of segregation. However, some provinces in the islands of Java and Nusa Tenggara (Central Java, Yogyakarta, East Java, Bali and East Nusa Tenggara) have a segregation rate of only half of the three highest provinces in Indonesia.
On Java, the relatively stagnant D index number also occurred in the provinces of DKI Jakarta, Central Java, East Java, and Banten, but the downward trend occurred in West Java

Discussion
Occupational segregation based on gender is caused by males and females doing different jobs and tasks in the labor market (Chapman & Harding, 1985). Occupational segregation also refers to the unequal distribution of female and male workers across and in types of work. Segregation can be horizontal, with females and males concentrating across different sectors, industries, occupations, product types and business sizes; and vertically, with gender differences in positions of different status, managerial responsibilities, or promotion potential. This is reflected in Figures 3 and 4 where males are mostly represented in managerial jobs, plant and machine operators, and other basic occupations. However, females are mostly concentrated on support workers. Figure 3 shows the percentage of females in different occupations from 2015 to 2018. In this figure, occupations are grouped in detail into large categories such as Sales Worker; Agriculture, forestry, hunting and fishing workers and laborers; and Production workers, operation of machinery workers. Horizontal segregation often puts females disproportionately into work in occupations that emphasize nonmanual skills. On average, 10 percent of females work as Agriculture, forestry, hunting and fishing workers and laborers, but only about 1 percent work in Managerial and supervisory occupations. Horizontal segregation also occurs if, within the main groups such Professional, technician and related occupations, females are more likely to work in occupations that are "people-oriented" than "object-oriented". Thus, in Indonesia, there are contrasts based on these two things. However, because the percentage of females who are included in the second criteria is only one-tenth of the first criteria, it can be said that there is no horizontal segregation.
A slightly different pattern occurs in male workers who work in leadership and management positions (Figure 4). There is about 2 percent of male workers in this occupation and three times the number of workers who work in occupations of Professional, technician and related occupations.
The occupations that are mostly carried out by female and male workers are the same, namely Agriculture, forestry, hunting and fishing workers and laborers. This is not surprising because the absorption of labor in Indonesia is still dominated by the agricultural sector, and the contribution of national income from the production side is also still the dominant contribution of this sector.
Significant differences occur in the following matters: (1) the occupation of the sales workers, between 20 and 25 percent of female workers work in this occupation, while only 10 to 14 percent of male workers. This is possible because females are more likely to work in occupations that are "people-oriented" than "object-oriented"; (2) the occupation of Production workers, operation of machinery workers is carried out by one-third of male workers while only one-fifth of female workers. This fact also confirms that more men work in occupations object-oriented.
The labor market according to the occupation, based on Table 3, is the most segregated in the occupation of production workers, operation of machinery workers. However, the trend has decreased in those four years. This is quite logical because this type of work requires physical strength, which is more male workers than female workers. According to this measure of occupational segregation, more than a third of women or men have to change jobs so that there can be true gender parity across jobs. Not much different from the findings of (Hegewisch & Hartmann, 2014) and (Weeden, Newhart, and Gelbgiser, 2018), which state that nearly half of females in the workforce have to change different jobs to eliminate all occupational segregation according to gender.
Changes in index scores can reflect changes in the composition of jobs or changes in the workforce mix of jobs that are more or less integrated among all occupations (Blau & Hendricks, 1979). Assuming all other variables are constant, the index decreases when more females move to jobs where they are underrepresented, such as in construction.The occupational segregation indicator makes it possible to determine the percentage of workers who have to be transferred to different sectors of the economy, sections or occupations in order to eliminate occupational segregation based on gender. The index is based on occupational segregation as the difference in employment between males and females in a given region. The lower the observed difference (the index value is close to zero), and hence, the more evenly the distribution of the labor force is, the smaller the scale of occupational segregation (Domagala, 2018) Gender-based occupational segregation between black female and male workers was much higher than among white female and male workers during the first two decades, but gender segregation indexes have converged for these two racial groups over the last decade. Overall, occupational segregation by gender is much stronger than occupational segregation based on race .
To examine the occupational segregation in Indonesia it can also be done in the context of the region (province). By calculating the Duncan index, each province can be assessed the level of horizontal segregation based on occupation as shown in Table 4. Bali, East Nusa Tenggara and West Kalimantan provinces are the three most integrated regions with an average index value of 0.1429 each; 0.1436 and 0.1883. On the other hand, the three most segregated regions are Bengkulu, North Sulawesi and Gorontalo provinces, namely 0.4095; 0.4033 and 0.3526.
The trend of increasing segregation during this period was shown by almost all provinces on the island of Sumatra, namely Aceh, West Sumatra, Jambi, South Sumatra, Lampung and Bangka Belitung. Generally, all provinces in Java show a fluctuating index. Other areas in Indonesia also show fluctuations in the phenomenon of occupational segregation, except for Papua. So, it can be stated that more than half of the provinces in Sumatra are increasingly segregated, while three provinces have stagnated in the D index (North Sumatra, Riau and Riau Islands) and Bengkulu province has increased sharply in 2016 and has decreased in the last two years.

Conclusions
The occupational structure in Indonesia shows that there is no occupational segregation based on gender where the index number D tends to be closer to zero. The occupation of the leadership and management of males and females workers is the most integrated. However, a trend segregation in this occupation from 2015 to 2018 increased as a whole. The same thing also happened to the service business occupation. According to the occupation, the labor market is the most segregated in the occupation of Production workers, operation of machinery workers. Integration is showing signs of increasing in Professional, technician and related occupations; Clerical and related occupations; Sales Workers; and Agriculture, forestry, hunting and fishing workers and laborers. In the regional context (province), Bali, East Nusa Tenggara, and West Kalimantan are the three most integrated regions, while the three most segregated regions are Bengkulu, North Sulawesi, and Gorontalo provinces.
The results of this research were conducted by examining the level of segregation based on gender and occupation of workers in Indonesia in almost all provinces. So that a more detailed picture can be obtained regarding the level of segregation to the provincial level. However, there are still limitations to this research, namely using only one approach to measure the dissimilarity (D-Index).