Indonesia has become one of the biggest migrant-sending countries in Asia after Sri Lanka and the Philippines (Hugo, 2009). But, are we aware of how many Indonesians are working abroad? No one probably knows the exact answer.
Estimates of the total number of migrant workers vary from 4 million (The Jakarta Post, 2012), 4.1 million (Bank Indonesia, 2011), to 6 million (Kompas, 2011). Meanwhile, the National Agency for the Placement and Protection of Indonesian Migrant Workers (BNP2TKI) only reports very outdated figures of migrant flows for August 2007 and the Manpower and Transmigration Ministry’s website reports deployments from 2007 to 2009.
Is it so difficult to generate timely and accurate statistics of migrant workers? Before coming to that discussion, it is worth reviewing the institutions responsible for producing information on migrant workers.
First and foremost is BNP2TKI. With 19 provincial representatives (BP3TKI) and 14 representatives at the regency level (P4TKI), BNP2TKI should be able to provide monthly data on officially deployed migrant workers. It records broader information collected at the point of departure and the point of entry back into Indonesia.
Second is the Manpower and Transmigration Ministry, whose information comes from private recruitment agencies (PPTKIS). Article 14 of Law No. 39/2004 on the protection and placement of migrant workers grants the ministry the power to request reports from PPTKIS at regular intervals prior to approval of PPTKIS permit extensions.
Third, the Transportation Ministry has data on workers aboard foreign commercial ships, but irregularly shares the data.
Fourth, the Law and Human Rights Ministry, through the immigration office, retains individual records of the workers since it differentiates migrants’ passports from common ones.
Fifth, the Foreign Affairs Ministry by means of the Indonesian embassies and consular offices also store such data.
Sixth, the Communication and Information Ministry has been preparing an online information system for migrant workers since 2008. Since the system totally relies on the data contributed by the relevant institutions, it has not progressed well.
According to Law No. 39/2004, the institutions in charge of keeping the information on workers are PPTKIS, local governments through registration (Articles 22 and 36) and Indonesian embassies (Articles 71 and 74). The law does mention information systems, but it is unclear about the integration of data collected by various institutions.
It should come as no surprise that the existing statistics have no information on the migrants’ provinces and districts of origin. This is because the data is only disaggregated based on the placement unit.
For example, 2007 data only ranked Bandung as the seventh highest placement unit below Jakarta; Nunukan, East Kalimantan; Surabaya, East Java; Mataram, West Nusa Tenggara (NTB); Tanjung Pinang, Riau; and Semarang, Central Java.
Jakarta — home to 90 percent of PPTKIS — has continuously been the highest contributor to migrant worker outflows from 2007 to 2009. It serves as the transit area particularly for workers heading to Saudi Arabia. Workers from all over Indonesia come to Jakarta to undergo training and obtain work certificates before applying for Saudi visas.
Thus, West Java workers might be counted in Jakarta, which would explain the low numbers from Bandung’s placement unit. However, workers from NTB, whose popular destination is also Saudi Arabia, might be counted twice – once in NTB and once in Jakarta. The same case of double-counting might also happen in Nunukan and Tanjung Pinang, which appear to be the biggest placement units for people en route to Malaysia.
Moreover, the recording system contains vertical and horizontal discrepancies across institutions. The data recorded at the national level by BNP2TKI is not necessarily the same as that at the provincial and district level.
For example, East Java deployment reported differently by BP3TKI and BNP2TKI. Data produced by BP3TKI is also not necessarily the same as that in the regency level. Finally, data collected by BNP2TKI and the Immigration Office should be identical in theory, but in reality could be quite different.
Aside from ministry data, Statistics Indonesia maintains accessible data on emigration figures. Since 2005, the Village Potential Census (Podes) includes a single question on the number of workers abroad. However, since it is generated from interviews with village heads, rather than hard evidence, village heads might confuse the question about migrant workers with the flow of emigrants.
This might help to clarify why the total number of migrant workers in the Podes 2005 and 2008 was only one-fourth to one-sixth of earlier estimates mentioned earlier. Podes should, therefore, include both the questions on the numbers and flow of workers.
We can conclude that many government institutions collect migrant workers data, but each of them is partial, incomplete, and most probably overlapping, resulting in double-counting. At the same time, workers migrating for a second time, those working independently, and professionals would be less likely to be included, resulting in undercounting.
Another inherent problem concerns the prevalence of irregular/undocumented workers where policies in the destination country play an important role. For example, Malaysia allows migrants to buy work permits upon arrival.
Thus, workers leaving the country with tourist visas, who are considered being irregular/undocumented by Indonesia are indeed regular/documented by Malaysia. Hence, the information based only on regular/documented workers might distort the overall picture of overseas employment.
Tirtosudarmo (2007) wrote that the number of irregular/undocumented migrant workers might approach two to four times the number of regular/documented migrants.
Ultimately, statistics are not the goal but rather the means and tools for better management of international migration. Improving the governance of overseas workers encompasses many aspects, one of which is the improvement of statistical monitoring. Limitations of emigration statistics and their public access have imposed major constraints on research efforts.
We recommend that Statistics Indonesia conduct a nationally representative household survey covering relevant information on emigration, including workers’ profile such as gender, age, address, profession, education, wage, and destination country. The survey should incorporate workers preparing to depart, those currently working, and those who have already finished their work contract.
The writer is a researcher at the SMERU Research Institute. The opinions expressed are her own.