Big Data & Labor Markets - Taking First Steps

Labor Market Information Systems (LMIS) have three important functions:

• They help public- and private-sector actors design labor market-oriented vocational training.
• They assist job seekers in finding vacancies, possible career paths, and suitable training opportunities.
• They help companies design their systems for human resources management.

However, in many partner countries of development cooperation, LMIS are patchy or dysfunctional for several reasons.

High Cost – The total cost of ownership cannot be borne due to high development and maintenance costs. Donor organizations are reluctant to make corresponding investments due to the complexity and costs involved.

Insufficient Resources – LMIS are only useful if they are constantly updated. In most cases, the responsible institutions in partner countries do not have the capacity or resources to run such systems sustainably.

Limited Scope – Often, LMIS are neither integrated nor cross-sectoral structured and therefore do not provide a comprehensive view of the labor market and its dynamics.

The digital transformation enables the utilization of non-traditional data sources to enrich LMIS. Big Data (e.g. of online job portals) can help analyze labor markets, creating a more comprehensive data picture. With a wider view of the data, information asymmetries (matching) and matching problems (linking) can be addressed.

Additionally, labor supply and demand can be captured more comprehensively and dynamically. This would enable job seekers to be trained according to labor market demands. Employers could also adapt their job advertisements and Technical Vocational Education and Training (TVET) programs. Policies could be designed with flexibility, and demand-driven employment services could also be more flexible and effective.

Indonesia, with its large and dynamic economy, has strong partner interest in improving its LMIS, a multi-stakeholder ecosystem dealing in labor market data with a strong presence of GIZ in the area of employment promotion and TVET. Indonesia was chosen as a pilot country to test the application of Big Data for LMIS purposes. GIZ Data Lab supported the initiative with a design workshop in April 2019 in Jakarta. The workshop brought together policy makers, data providers, and technical experts from education, professional, and institutional backgrounds related to labor markets from the private and public sectors.

The workshop participants identified their preferred project idea, One Data, and conducted an evaluation and feasibility study. The government of Indonesia is already implementing a One Data program where different government datasets are aggregated into a central portal for public access.

The GIZ idea aims to contribute to the innovative use of Big Data in the context of the existing governmental efforts by strengthening labor statistics and ensuring that high-value datasets on labor markets are utilized. Moreover, the project will build on current data initiatives of agencies concerned with and mandated to strengthen labor market development in the country.

One Data will make use of official statistics, agency-specific datasets (e.g. manpower, education) and alternative data sources (e.g. job portals, data aggregators, secondary big data) in order to enrich current public datasets with data collected from the private sector. It requires a process of data selection and prioritization (e.g. what datasets are needed by different sectors?), data publication (e.g. how will the government disclose the data and in what format?), and data sharing (e.g. how do we incentivize the private sector to share data of public interest?).

In terms of next steps, the following process was proposed:

• Identify a pilot sector for the One LMIS Data experiment. Taking available resources, feasibility, and time constraints into account, the tourism sector is the preferred option. As an alternative, the manufacturing sector offers similar opportunities.
• Align the concept with what is currently being proposed by other donors and agencies so that an early experiment is aligned with the strategic direction of the Indonesian government.
• Secure political commitment and buy-in not only from the government, but also from key players in the private sector.
• Develop a concept note including an estimation of required resources and then raise funding for the initiative.

The great interest of the partners and the increasing availability of decentralized data suggest great potential for further development.