What can online job postings say about the labor market dynamics in Singapore?

Online job postings complement official labor market data well and, therefore, reflect the skills and qualifications in demand in a timelier manner. A study by ACI uses job posting data from JobTech, Singapore’s leading AI talent intelligence startup, to monitor and understand local labor market dynamics. The data constitute nearly the entire universe of job ads in Singapore from 2018 to 2021. It includes the full text of the job descriptions. The richness of the data allows for an in-depth study of Singapore’s labor market. Therefore, the findings of this study fall under various themes discussed below.

The first finding allays concerns about the ability of job data to track the small Singapore urban labor market. It is found that job postings follow the labor market well (as seen in Figure 1). This means that the job data used represented the actual job market in Singapore quite well. When split by broad occupation groups, job postings could not track some groups, such as services and construction Second, the share of jobs not requiring formal diplomas and degrees was on the rise in 2021 relative to previous years. A closer look into the salaries, however, revealed a different story. There was no significant change in the salaries of such jobs over the years. This means that their increasing share “reflects a compositional change in the economy rather than a change in hiring practices.”

Third, focusing on the highly demanded skills, it was found that those related to STEM (Science, Technology, Engineering, and Mathematics) fields saw the most significant increase in their demand over the years. Skills relating to scripting languages (Python) and programming languages (Java) recorded the fastest growth. Next, delving deeper into AI-related job postings, it was observed that they have been on the rise since at least 2020. Moreover, the stated salary in job postings is higher than those in non-AI postings. This indicates the importance and, perhaps, a lower supply of AI-related skilled candidates.

The onset of the Covid-19 pandemic made work-from-home (WFH) or Flexi-work arrangements more commonplace, if not default, workplace setting. This change, however, was more concentrated in some sectors, such as the services sector. It was found that WFH jobs required ICT skills such as cloud technology, amazon web services. On the other hand, non-WFH jobs require more ad-hoc duties, such as customer support and maintenance management. Therefore, the opportunity to WFH was not available equally to everyone in the workforce.

The last finding is based on the observed differences between the government-run job board, MyCareersFuture, and four private-run job boards. It is found that the former was much more likely to state salary in their job posts as compared to their private counterparts. Further, the educational attainment requirements on MyCareersFuture tend to be relatively higher. This, the authors argue, could be “due to differing expectations from employers when advertising for jobs on different online portals and the self-selection of job seekers as well.”

Assessing the labor market using JobTech’s data sheds light on prevalent job requirements, such as skills and qualifications. The data suggest that STEM skills, especially programming and machine learning, are in high demand among employers. The findings could complement policymakers’ efforts in preparing their workforce to thrive in an increasingly digitized economy. Moreover, with the improving Covid-19 situation, knowing to what extent WFH will remain a permanent fixture as employers increasingly restore normal working conditions is useful. This has implications for urban planning, including the transport system.

By GUPTA, Shubhangi

Researchers: HO, Wai Luen Terence, LIM, Jing Zhi, SHEN Yan Shun, Lucas

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s