Recent Survey on Indonesia Employment: A Brief Analysis

This article is aimed to analyze briefly the recent trends of Indonesia’s working-age population and its components as viewed from the labour force perspective. The analysis covered the period 2008-2014 and focused at the national level of analysis. Sakernas– a regular national labour survey carried out by BPS-Statistics Indonesia since 1976– is used as the major data source. In order to have a better understanding of the measurements used in the analysis, the following paragraphs discuss briefly some conceptual issues concerning some basic labour statistics.

Labour Statistics: Global Standard

Sakernas adopts the global standards of labour statistics as stipulated in the resolutions of the International Conference of Labour Statisticians (ICLS). In this context, the 13th ICLS (1982) and the 19th ICLS (2013) are of special interest as they set out the global standards for basic[1] labour statistics such as employment and unemployment.

With regards to persons in employment, Sakernas defines it as all those who during the last week, were engaged in any activity to produce goods or provide services for pay or profit. This definition is fully in line with the global standard (ICLS-19, Par. 27). Sakernas defines persons in labour force as those in employment and in employment and this also complying the global standard (Par. 16).

What might worth noting here is that official statistics derived from by far has not yet taken into consideration those who engaged in productive activity not for employment but for “own use” and for “volunteer”. According to the 19th ICLS, these two categories are not part of employment[2].

By so doing, official employment statistics produced by BPS by far might be regarded as somewhat overestimate. Nonetheless, that is not an issue for this analysis as it focuses on the trends that require consistency in applying the concept throughout the compared period.

Demographic Dimension

One of the biggest challenges faced by Indonesia is a sheer number of the working-age population (WAP). This is not surprising as the country ranks the fourth biggest country after China, India, and the United States. To add the challenge, as Sakernas series data show, the WAP tends to increase in a pace that is faster than the growth rate of the total population. WAP data for the period 2008-2016 can be used to illustrate the increase. During the period total WAP increased from around 174.2 million in 2010 to 189.8 million in 2016. This is an annual increase of 1.86%[3]. This pace of increase is much faster than the increase in the total population during the same period that was 1.36%.

The comparison of the figures points to a phenomenon popularly known as the demographic bonus. That can be a bonus– and not a curse– if the younger generation is able to get a decent education and facility to improve their self-quality[4].

There is another aspect of the phenomenon just mentioned worth considering. Close observation at the trends in the WAP by age group provides a strong indication that the status of demographic bonus for Indonesia is at a somewhat later stage. The following points might be helpful in clarifying the concerned issue:

    • During the 2008-2018 period (August) the increase in the proportion for the younger-ages group (15-24) was very small and even almost flat; that was, only 0.35%.
    • The percentage was higher for the “prime-age” group (25-54) that was 1.43 %.
    • For “old-age” group (55+) the percentage was even much bigger; that was 3.7% (See Graph 1).

Graph 1: Working-age Population by Age Group, 2008-2012

Source: BPS, Sakernas,

The comparison of the increases confirms the above notion of the status of a “later stage” of demographic bonus of Indonesia. In addition, the relatively high increase for the old-age group suggests a clear indication of the route of Indonesia’s population toward the aging stage.

Trends in the Labour Force

As shown by Graph 2, the proportion of “out of the labour force” remained unchanged during the period 2008-2012; that was, 33% of the total WAP. This means the labour force also unchanged at 67% level during the same period.

In the same period, the compositions of the labour force had changed by three percentage points but with different direction: while the employment increased by three percentage points, the unemployment decreased by the same percentage points[5]. (See Graph 2.)

Graph 2: Change in the Structure of Working-age Population 2008-2012.

Source: BPS, Sakernas,

Graph 3 shows the increase in the labour force during the period 2008-2018. As shown by the graph, the pace of the increase is not as fast as the increase in the WAP. As also shown by the graph, the labour force was almost always bigger in February than in August and this is probably associated with the seasonal work in agriculture, a still important economic branch for Indonesia’s employment.

Graph 3: Trends in the Working Age Population and Labour Force (000)


Source: BPS, Sakernas,

Trend LFPR and EPR

Statistics of employment can be measured by two indicators of the labour force participation rate (LFPR) and the employment-population ratio (EPR). These two indicators are comparable as each of them using the same numerator or “the population at risk” that is the WAP[6]. Nonetheless, each of them is to serve its own function. While the first indicator reflects the supply side of the labour market, the second reflects its demand side; i.e., measuring how much the labour supply absorbed by the economy.

Graph 4 exhibits the comparison between the LFPR and the EPR during the period 2008-2018: the LFPR is always higher than the EPR for the obvious reason: the numerator of the LFPR includes unemployment element that is not included in EPR.

Graph 4: Trends in the Labour Participation Rate and Employment-Population Ratio


Source: BPS, Sakernas,

The graph shows the LFPR was around two-thirds of the WAP and tends to be higher in February than in August for the reason as previously mentioned. The graph also shows that the LFPR tends to fluctuate over the observed periods with slightly different direction: the trend slightly increasing in February and increasing in August.

Like LFPR, EPR is higher in February than in August. Unlike LFPR, EPR tends to increase over the compared periods. This suggests the consistency in the increase of employment regardless of the month of observation (February of August).

Trends in Unemployment

During the period 2008-2018 the unemployment rate in Indonesia was relatively low (one-digit) and tends to decline. As shown by Graph 5, during the period the unemployment rate declined from around 8.4-8.5% in 2008 to 5.1-5.3% in 2018. Comparison between levels in February and August shows that during 2008-2013 the unemployment rate tends to be higher in February (than in August), and starting 2013 the reverse was in place; i.e., tends to be higher in August (than in February).

Graph 5: Trends in the unemployment rate (%), 2008-2018

Source: BPS, Sakernas,

The relatively low and decreasing rates in unemployment as just discussed do in fact obscure a high unemployment rate among the educated persons: on the average, the educated persons have 2.6 bigger risks of being unemployed than the uneducated. (See this for detail.)

Concluding Remarks.

Conventional wisdom suggests that the age-structure of Indonesia’s population in the 2020s will be entering the phase of demographic bonus. Quite contrary to this wisdom, a series of Sakernas data do indicate the phase has been entered since the 2010s and even in its later phase. The evidence of this is that the pace of increase in the working-age population than in the total population and this especially striking for the older-age group (55 or older).

The LFPR remained constant during the period 2008-2018 at the two-thirds of the total WAP. However, these changes could be misleading as the trends in its components went through in a different direction; that was, the increase in the employment and the decreasing in unemployment.

The unemployment rate during the period 2008-2018 was at relatively low (one-digit) and its trend decreasing continuously. However, this obscures the high level of unemployment rate among the youth. This issue– together with other issues like the forced labour and “modern slavery”– is probably one among the undesirable modern paradoxes. Who knows?

[For pdf version click this.]


[1] Statistics on such issues as SDG indicators and child labor are not regarded as basic statistics and hence not covered in the analysis.

[2] Starting 2016 the Sakernas questionnaire is refined by additional questions that can provide data on these two categories. This refinement can also provide much richer data on all forms of work that have been identified by the 19th ICLS.



[5] Expressed in a different way, during the period, the total employment increased by 21.5 million or 21.0% while the total unemployment decreased by 2.4 million or 25.5%.

[6] LFPR = labour fore/WAP and EPR = employment/WAP. The complement of the second reflects the magnitude of the unemployed over the WAP.



Labour Underutilization: Concept and Measurement (3/3)

Section 3: Some Lessons from the 2012 Sakernas

As discussed above, the Sakernas (until 2015) is unable to provide data on “potential labour force” (PLF), the second component of the labour underutilized” (LU). The reason for this is that the questionnaire of the survey does not contain the question of “the availability of work”. However, part of PLF (using ICLS-19 standard) has been included already in the unemployment rate; namely, “discouraged job seeker” (DJS).

The estimated population of DJS, according to Sakernas 2012, is around 2.26 million. This is a big number as reflected in the DRS-unemployment ratio which is about 31:100. The table also shows some numbers that can be used to estimate population or ratios of some components of Working-age Population (WAP) as below:

  • Unemployment (U) = 7.2 million, if DJS is considered as part of U (as official figure); OR 5.0 million, if DJS if DJS is considered as part of “Outside LF)” (as suggested by ICLS-19);
  • Unemployment rate= 6.1% or 4.3%; depending on how to treat DJS;
  • Time-related Underemployment (TRU) = 11.5 million
  • Labour Underemployment (LU) >= 18.8 million, if LU>=U+TRU, and
  • LU rate >= 15.9%, if LU rate = (LU/Labour Force) *100.

[Beck to Section 1]

Labour Underutilization: Concept and Measurement (2/3)

Preliminary notes:

The concept of labour underutilization as discussed in the first section is the product of the resolution of  ICLS-19 hosted by ILO aimed mainly to be used as global guidelines in the area of labour statistics.  However, as a product of an ILO resolution, the concept is not binding for the participating countries of ILO. They might not be able to promptly follow the guidelines for practical reasons and hence need some time to apply in their actual survey. Part of the reasons for the participating countries are: (1) the need to “harmonize” the concept with actual situation of labour market they face, (2) the need to maintain “consistency” (as opposed to “validity”) of labour statistics between years to avoid confusion among data users, (3) the need test carefully the practicality of the concept in actual survey, and (4) the need to follow their own priorities in statistical activities. 


Section 2: Data Availability

At first glance, it appears to be that all the proposed components of “labour underutilization” as outlined in Section 1 are readily produced by a standard labour force survey. However, that is not fully the case, at least in the case of Indonesia. Here is a brief description on that issue.

Until 2015[1], the questionnaire of Sakernas, or Indonesia Labour Force Survey (ILFS) has no question on “the availability of work”. (What is available is a question on “the readiness to accept an offer for more job” that is intended to capture “time-related underemployment” as discussed in Section 1.)

In order to produce official statistics on unemployment, Sakernas defines unemployment put simply as:

(“Not in employment”) & ((“Seeking work”) OR (“Not seeking work due “Future start”” OR “Discouraged”)).

The above definition results in the figure of unemployment as mentioned in Section 1. The definition clearly shows that “availability for work” is ignored in defining “unemployment”.

The above definition “correctly” includes “future start” (not seeking because of having a job already) but “wrongly” includes “discouraged” (not seeking because of feeling there is no opportunity) in the unemployment. According to ICLS-19, “discourage” job seeker belongs to “outside labour fore” category, not “unemployment” (hence not in “labour force” category). In other word the official statistics of unemployment of Indonesia has in fact already included “discouraged” component of “outside labour force” (per ICLS-19 standard). As will be shown soon in Section 3, this component is relatively big, roughly 31% of the unemployment in 2012.

[1] Since 2016 BPS has initiated to improve the Sakernas questionnaire in order to address most of the issues raised by ICLS-19. Processes to refine the questionnaire toward this direction are still in place, until now.

[Proceed to Section 3: Sekernas’ Lessons]


Source: Google


Labour Underutilization: Concept and Measurement (1/3)

Section 1: Unemployment and Labour Underutilization

According to Sakernas 2018 or the 2018 Indonesia’s Labor Force Survey, the estimated total of the working age population (WAP) of Indonesia (2018 ILFS) is around 194.8 million. Out of the total, 131 million are classified as labor force (LF) and seven million are the unemployed. The unemployment rate is then about 5.3%.

While many might view the unemployment rate is comparatively low, it is basically sensible given these two facts:

    • Around two-thirds of Indonesia’s employment are engaging in the informal sector, and
    • There are no social security systems applied for unemployment in this country.

In addition, given the big population of Indonesia, even such a low unemployment rate equivalent with 1.2 total population of Singapore. For further discussion on this see THIS.

Perhaps only a few (if any) who disagree on the importance of the statistics of the unemployment rate as is a leading indicator for labour market. Likewise, perhaps only a few who disagree with the notion that the unemployment rate alone already reflects the situation of the labour market appropriately.

Many believe that changes in the unemployment rate is an insensitive indicator to track the real situation in the labor market. The economic crisis, for example, this indicator does not provide a clear signal to policymakers to anticipate. In the case of the 1997 Indonesian crisis, as another example, “a puzzle” was even found: during the Mid 1997 period (before the crisis) and the end of 1998 (when peak of the crisis ended): the number of employment increased by about 1.5% and the number of unemployment decreased by 12.3% [1].

The question would be then what other statistical measures– beside unemployment rate– that can be used to reflect and monitor the dynamics of the labour market in clearer, more realistic, and more sensitive way. The 19th International Conference of Labour Statistician in 2013 (ICLS-19) accentuates such an issue and promote pose the concept of “labour underutilization” (LU).

ICLS-19 (Par. 40) proposes this concept that includes these three elements; namely, unemployment, time-related underemployment (TRU),  and potential labour force (PLF). Here are the definitions of them.

    1. Unemployment = (not in employment) AND ((seeking work AND available for work));
    2. TRU = (in employment) & ((worked less than a normal working hour) & (seeking and available for more job)).
    3. PLF (Pars. 51-55):
      • (not in employment) & ((seek empolyment) & (were not :currently available”)); i.e., unavailable job seekers, OR
      • (not in employment) & ((not “seek employument”) & (“currently available”)); i,e., available potential jobseekers.

These measures are

the basis to produce headline indicators for labour market monitoring. For more comprehensive assessment they can be used with other indicators relating to the labour market, …. in particular skill-related inadequate employment and income-related inadequate employment.. ” (Par. 41)

It is worth noting that while the first component mentioned above belongs to labour force, the third belongs to “outside labour force” as generally understood. A reference for ICLS-19 can be accessed HERE.

Graph 1 provides a schematic presentation of the components of LU as just mentioned. The graph shows among others that unemployment is only a fraction of a much larger LU category.

Graph 1: Composition of working age population

[1] Puguh Irawan and Uzair Suhaimi (1998:11) in Crisis, Poverty, and Human Development in Indonesia, BPS-UNDP.

[Proceed to Section 2: Data Availability]

Profil Ketenagakerjaan Indonesia Berdasarkan Survei Terkini

 Sumber Gambar: Google


Tulisan ini mengkaji secara singkat profil ketenagakerjaan Indonesia berdasarkan hasil survei terakhir yaitu Survei Angkatan Kerja 2018 (Sakernas 2018). Bagi Indonesia survei ini merupakan sumber statistik resmi (official statistics) dalam bidang ketenagakerjaan. Fokus kajian adalah komposisi penduduk usia kerja (PUK) dan salah satu komponennya yang utama yaitu pengangguran. Untuk memperoleh gambaran mengenai perkembangan antar waktu, Sakernas tahun-tahun sebelumnya juga digunakan.

Penduduk Usia Kerja

Profil ketenagakerjaan Indonesia dapat digambarkan secara singkat sebagai “serba besar”. Sebagai ilustrasi, penduduk usia kerja (PUK) menurut Sakernas 2018 berjumlah sekitar 194,8 juta jiwa. Besarnya angka ini sebenarnya wajar karena dengan total penduduk sekitar 268 juta jiwa Indonesia menempati urutan keempat negara terbesar setelah China (1.39 milyar), India (1.36 juta), Amerika Serikat (327 juta). Besarnya angka PUK Indonesia itu kira-kira setara dengan tiga kali angka keseluruhan total penduduk lima negara jiran terdekat yaitu Timor Leste, Australia, Singapura, Malaysia, Brunei[1].

PUK Indonesia == tiga kali angka keseluruhan total penduduk Timor Leste, Australia, Singapura, Malaysia, dan  Brunei

PUK dapat dibagi habis ke dalam tiga komponen utamanya yaitu “bekerja” (B), “penganggur” (P) dan “bukan angkatan kerja” (BAK). Masing-masing komponen ini eksklusif dalam arti tidak saling beririsan sehingga PUK = B + P + BAK. Istilah Angkatan Kerja (AK) merujuk pada gabungan B dan P.

Dalam persamaan ini diberlakukan aturan prioritas: B terhadap komponen lainnya dan AK terhadap BAK. Dengan aturan ini ada kepastian mengategorikan status ketenagakerjaan setiap responden survei.

Berapa besar B dan P? Sakernas 2018 menujukan angka total masing-masing lumayan besar: yaitu 124 juta dan 7 juta. Dengan demikian AK berjumlah sekitar 131 juta. Dari angka-angka ini dapat dihitung dua indikator ketenagakerjaan yaitu “angka penganggur” (AP) dan “rasio tenaga kerja/penduduk” (RTP):

    • AP = (P/AK)*100 = (7/131)*100 = 5.3%
    • RTP = (B/PUK)*100 = (124 /194.8) *100= 64%.

Sebaran umur dua indikator ini dapat dilihat pada Tabel 1. Pada tabel ini angka penganggur dapat diperoleh dengan mengurangi angka 100 dengan angka-angka pada kolom “%Bekerja/AK”.


Seperti terlihat dalam Tabel 1, angka penganggur di Indonesia adalah 5.3%, suatu angka tergolong kecil. Yang perlu dicatat, angka mutlak dari angka persentase yang kecil masih jutaan, 7.0 juta jiwa. Angka ini setara dengan 1.2 total penduduk Singapura[2].

Total penganggur Indonesia== 1.2 total penduduk Singapura.

Relatif kecilnya angka penganggur itu “menyembunyikan” permasalahan yang lebih struktural: angka penganggur yang didominasi oleh penduduk usia muda dan kelompok terdidik.

Sumber: INI

Tingginya angka penganngur untuk kelompok usia muda dapat dicermati pada Tabel 1. Perkembangannya antar tahun dapat dilihat pada Gambar 1.

Sumber: INI

Catatan: 2011-2013 hasil backast menggunakan penimbang perbaikan berdasarkan angka proyeksi penduduk.

Tabel 2 menunjukan relatif tingginya angka penganggur bagi kelompok terdidik (tamatan SLTA+). Menurut tabel itu risiko penganggur 2.6 kali lebih tinggi bagi kelompok terdidik dibandingkan dengan kelompok tak-terdidik.

Risiko pengaggur bagi penduduk terdidik == 2.6 kali risiko bagi yang non-terdidik.

Tingginya angka prevalensi penganggur bagi kelompok terdidik tercermin pada Gambar 2. Pada tahun 2018 terlihat, misalnya, dari 100 orang penganggur, 66-67 di anataranya terdidik. Gambatr itu juga mencerminkan bahwa tinginya angka itu bukan hal baru dan kecenderungannya memburuk.

Sumber/Catatan: Sama dengan Gambar 1.

Sebagai catatan, semua istilah, aturan, dan rumus penghitungan yang dikemukakan di atas mengacu pada standar global dalam bidang ketenagakerjaan. Standar ini tercantum dalam resolusi International Conference Labour Statistician yang ke-13 (1982) atau ICLS-13. Belakangan disadari adanya sejumlah permasalahan konseptual pada reolusi ICLS-13 ini. Permasalahan ini dicoba diatasi melalui resolusi ICLS-19 (2013) [3].

BPS merespons aspirasi ICLS-19 ini sejak 2016 sekalipun  sejauh ini baru pada tahapan penyempurnaan kuesioner. Upaya ini perlu diapresiasi dan didukung oleh pemakai data Sakernas. Alasannya, Sakernas berbasis ICLS-19 dapat diharapkan menghasilkan sejumlah headline indicators yang lebih lengkap untuk memotret profil, lebih cermat dalam mengukur besaran, serta lebih peka dan realistis dalam memantau dinamika ketenagakerjaan di Indonesia dalam terang standar global.


[1] Angka penduduk diambil dari SINI.

[2]  Lihat catatan kaki-1.

[3] Rujukan mengenai ICLS-19 dapat diakses di SINI.

19 Selected Tables from the ILFS

If you are interested in employment-related issues in Indonesia, this post is the right one for you.


Sakernas, or Indonesia Labour Force Survey (ILFS), provides regularly a number of tables concerning labour statistics of Indonesia. Some of the tables (in Excel) are published regularly HERE.

The 19 tables found that link is presented at the national level, disaggregated by age groups, gender, type of residence, and educational level. In some cases, time series data are available there.

HERE is the list of the tables.


Source: Google

International Conference of Labour Statisticians (ICLS): A Brief Note

Sumber gambar: Google


(i) Function and Participants of ICLS:

  • Global standard-setting mechanism in labour statistics
  • ILO hosts & acts as Secretariat
  • Meets every 5 years (since 1923)
  • Tripartite structure: Governments (NSO, MoL), Employers, and Workers representatives
  • Observers: International and regional organizations, NGOs


(ii) The objective of ICLS:

Main objectives of ICLS statistical standards

  • Provide guidance to countries in setting their national labour statistics programmes
  • Promote coherence in concepts & methods across sources & topics / areas
  • Promote international comparability, and
  • Set priorities for future work


(iii) The most recent ICLS:

The most recent ICLS (the 19th) took place in 2013. It produces, among others, “Resolution 1: Resolution concerning statistics of work, employment and labour underutilization’ that contains 97 Paragraphs.

The complete version of the resolution can be accessed Here; some excerpts, Here.