(Content warning: This commentary is centred on Singapore’s sexual violence and technology-facilitated harm problem, with references to sexual assault, rape, and sexual abuse.)

Let’s start with the reported numbers. Over the past five years, around 2,500 sexual assault cases are reported each year. Reported molestation cases increased from 1,427 in 2024 to 1,531 in 2025 (a 7.3 per cent increase). Similarly, over the same one-year period, the number of reported rape cases rose by 19.5 per cent, from 401 to 479. The majority of rape cases and more than half of outrage-of-modesty cases involved offenders known to the victims. Furthermore, sexual offences ranked second among offence categories for youths aged 10 to under 16.

These numbers should also be read together with Singaporeans’ exposure to harmful online content and behaviour. In its 2025 surveys, the Ministry of Digital Development and Information found that among those who had encountered harmful online content in the previous year, 26 per cent of them encountered sexual content at least a few times a month (second in frequency to content supporting illegal activity; e.g., scams, sale of prohibited items). Among respondents who had experienced harmful online behaviour, 27 per cent received unwanted sexual messages (second to catfishing). Although sexual content and messages are not necessarily sexual violence, these findings help contextualise a wider digital environment where unwanted sexual conduct occurs.

As the government and its agencies have strengthened prevention (e.g., school engagement, public education, awareness) and enforcement actions (e.g., higher penalties, covering AI-generated materials), recent research reports have also helpfully coalesced around diagnoses and recommendations. (As a quick aside, the topline figures I cited above were not broken down by sex or gender in the sources.) I summarised five common themes, after synthesising four of these reports:

1. Study on online harms in Singapore (2026). Involving 1,048 Singapore Citizens and Permanent Residents (aged 15 and above), the report studied the prevalence and impact of online harms, while examining avenues for help and proposing recourse and recommendations. It was published by SHE, SG Her Empowerment, a non-profit organisation.

2. Survey on men’s attitudes to gender, family and society in Singapore (August 2026). This report has a broader scope of understanding Singaporean men’s cultural expectations and attitudes, with relevant questions – and findings – on sexual entitlement beliefs as well as consent and accountability, among others. Advocacy organisation AWARE Singapore, with YouGov, surveyed 1,084 men in October 2025 for this study.

3. ‘When everything goes online, it’s never really gone’: understanding technology-facilitated sexual violence (TFSV) in Singapore (September 2025). The researchers analysed 2,994 valid survey responses from students at the National University of Singapore (NUS) and interviewed 20 students, finding a higher proportion of technology-facilitated sexual violence (TFSV) than in the comparison figures cited by the authors, as well as perceived normalisation of TFSV.

4. 404 Help Not Found: Lived experiences of online harms survivors (2025). Focused on understanding the impact of online harm, 25 survivors with such experiences were interviewed. In particular, in the aftermath of their experiences, survivors described feeling unheard by the platforms, while police reports and other formal processes could feel intimidating or disproportionate. The report was also published by SHE.

The AWARE Singapore report briefly referenced findings vis-à-vis gender roles in a King’s College London and Ipsos International Women’s Day 2026 study. For example, 64 per cent of Singaporean respondents said that equal rights for women had “gone far enough.” The proportion was 70 per cent among men, and 59 per cent among women.

Methodologically, bear in mind that the four reports use different methods, each with its own limitations. The 2026 SHE report used online surveys, the AWARE report surveyed men on their attitudes, the TFSV study used mixed methods, while the “404 Help Not Found” report involved qualitative interviews with survivors. Collectively, my focus was not on combining prevalence estimates but identifying common themes.

Theme 1: The normalisation of sexual violence

This thread of normalisation runs tightly across the studies. First, in the AWARE report, 37 per cent agreed that men cannot be expected to stop sexual activity if a partner changes her mind, and more than one-third believed that men “cannot control themselves” when sexually aroused.

Second, in the study involving NUS students, the interview participants described TFSV – defined as “a range of behaviours aided by digital technologies to harm, offend, humiliate, or intimidate individuals” – as routine, inescapable, or expected. From the survey, victim-survivors were somewhat more likely than respondents without TFSV experience to perceive TFSV as expected, but the overall average for both groups remained low. The researchers concluded (all pseudonyms):

“Similarly, our interviewees perceived TFSV as routine or expected and described TFSV as inescapable and ‘never really gone’, and its abuse in terms of a vicious cycle. Mel, Finn and Sienna appeared unsurprised not only at how pervasive TFSV was, but also the ease with which it could be perpetrated. That Marni, Vivian, and Yelena readily retold similar experiences of those around them only reinforces how ingrained such attitudes towards TFSV were within the society.”

And third, in SHE’s “404 Help Not Found” report, blame and normalisation – including and especially online sexual harassment against women – were characterised as one of its four recurring “fractures.” For example, one of the survivors reflected:

“It’s an unspoken understanding that if you’re a hot girl [or] just anybody mildly attractive… [you] must be ready for anything creepy… It’s the same from all my friends that I have spoken with.”

Critically, even though women are disproportionately affected by several forms of sexual and technology-facilitated harm, men can also be victim-survivors. A few of the reports touch on this. Moreover, drawing from broader research, the TFSV study indicated that LGBTQ+ people are also targeted due to their gender and/or sexual orientation.

Share

Theme 2: Technology changes the scale, speed, and permanence of harm

Against that backdrop of normalisation and existing gender norms in Singapore, technological advances – including social media and AI – are linked to a greater scale and scope of sexual violence. The harms to victims are more persistent, while enforcement is probably more challenging (e.g. anonymous accounts and difficulty identifying responsible individuals, cross-platform circulation, delayed takedowns). In other words, technology did not create sexual violence, yet it can make it worse.

Both SHE reports corroborate these observations. In its 2026 report, which concerns online harms generally, SHE documented the continued prevalence of such harms, with about a quarter of respondents personally experiencing at least one such harm in the past year.

More specifically, the report found that those dealing with Intimate Image Abuse (IIA; i.e., distributing someone’s nude, intimate, or sexually explicit media without consent) and Statements Affecting Reputation (SAR; i.e., statements, images, or videos harmful to a person’s reputation) had experienced these harms at a higher frequency, even if, overall, they were less common than other online harm categories. The report also distinguished between authentic and inauthentic IIA, with AI tools making the latter easier to generate.

In terms of the circulation of IIA and SAR, social media platforms continue to play an important distribution role. In the “404 Help Not Found” report, a survivor, Jack, was threatened with an AI-generated sexual video purporting to depict him. When the video spread, he received multiple Instagram direct messages demeaning and ridiculing him, with no way to stop its circulation or continued viewership:

“My parents… found a leaked video of me performing [a] sexual act. Then I just took a look at the video and some of them my face is topped with a video of the guy. Then that’s when I noticed how grave the matter is… it’s the aftermath… a lot more people… step up and text me on the Instagram, ask me, they start calling me names. Say that I’m ugly or what but still dare to post this kind of sexual video… a lot of names calling going on.”

Theme 3: Online sexual harm does not stay online

It is increasingly difficult to separate online from offline sexual violence, both in form and consequence. The “404 Help Not Found” report that nearly half of the 25 online-harm survivors interviewed by SHE had been harmed by someone they knew. Separately, in the 2025 police data presented earlier, the majority of reported rape cases and more than half of reported molestation cases involved known offenders.

Online harms have real-world impacts. The same SHE report documented the extensive ramifications, including professional and academic disruptions to work and/or school, spillover effects on their personal and social relationships (e.g., not being able to connect or be present with others), as well as strains on their personal well-being (e.g., poor sleep, appetite loss, anxiety and depression). Relatedly, after emphasising the permanence and uncontrollability of TFSV and its harms, researchers in the NUS study argued: “TFSV’s harms are therefore not limited to the act, but could live on in time and space, creating feelings of helplessness in victim-survivors.”

Theme 4: Egalitarian ideals coexist with traditional gender scripts

Even as technology changes the form and reach of sexual violence and harm, older gender scripts persist. Recall the earlier King’s College London and Ipsos International Women’s Day 2026 study, which found that Singaporean men and women differed in whether they believed equal rights for women had “gone far enough.” Also consider an earlier commentary of mine, about how care work remains gendered (e.g., married women continuing to do more child-related care work at home compared to their husbands).

Take a look at the AWARE report. Substantial majorities agree that childcare and housework should be equally shared, but perceptions don’t always translate into action (or the actual distribution of household responsibilities). It should also be noted that the AWARE report did not measure observed household behaviour. Why, then, do I feel that potential progress may remain slow? The same report found that “younger men aged 18 to 24 were the most likely of all age groups to have a stronger endorsement of traditional divisions in childcare,” though gender attitudes appear domain-specific and uneven (e.g., younger men in this sample were not uniformly more egalitarian).

It’s critical for me to add that sexual violence does not automatically result or follow from unequal caregiving beliefs. I argue that understanding the broader societal context, including the uneven changes in gender attitudes, may help make sense of the next steps. Which leads me to my final theme.

Theme 5: Survivors still carry too much of the burden

That survivors disproportionately bear the burden of sexual violence and its impact – which takes the form of victim-blaming (e.g., blame for having flirted, dressed a certain way, shared an image), self-protection (e.g., responsibility for blocking and repeatedly seeking takedown), difficulty identifying perpetrators, and the burden of deciding whether to pursue police or legal processes – sits uneasily at the intersection of prevailing gender attitudes and how technological platforms are designed.

Take victim-blaming attitudes for a start. In the AWARE report, 55 per cent of its male respondents believed that a woman who dresses provocatively is partly to blame if she is harassed, and 67 per cent thought women who flirt bear at least some responsibility for what follows. In terms of help-seeking and platform processes, the 2026 SHE survey reported that 43 per cent of survivors who did not report to platforms anticipated inaction or delay. SHE’s “404 Help Not Found” report also offers qualitative evidence, in terms of blame:

“I’m just scared that I’m not taken seriously or reporting does no harm [to the perpetrator]… that they will question me instead… that they would think that I’m part of it, you asked for it… what if they think [that] it’s all consensual or it’s all fun and games until I decided it’s not.”

The report also documented uncertainty when survivors engaged with tech platforms, including when seeking content removal or reporting users:

“Even if I report it… there’s bigger fish to fry. I don’t think I’m the worst case. I don’t think they’ll pay any attention to me.”

Subscribe now

What would shift the burden?

Every report I referenced here offered a list of practice and policy recommendations. The government has also been active. Senior Minister of State Sim Ann, at the annual Sexual Crime Awareness Seminar, described the harm and severity of sexual crimes and explained how the Ministry of Home Affairs has strengthened laws and enforcement. She also emphasised the protection of victims: “Victims are also encouraged to step forward and report a crime regardless of when it happened. Doing so as soon as possible can help preserve evidence, prevent further harm and bring a perpetrator to justice.”

Since June this year, victims of specified online harms in Singapore – including online sexual harassment and intimate-image abuse – have been able to turn to the OSC, or the Online Safety Commission. The OSC may issue directions disabling access to harmful material and restrict responsible accounts. The Commission received over 500 reports in its first two months, resolving more than90 per cent of the reports it deemed eligible.

A 2025 working paper by the Institute of Policy Studies, which included a 600-respondent survey, also found widespread support for laws holding perpetrators accountable (79.3 per cent) and faster platform takedown (77 per cent). Around three-quarters also thought that the government, technology companies, and individuals should do much more to improve online safety.

Be that as it may, especially with the rapid proliferation of AI, perhaps we ought to tackle Singapore’s technology-facilitated sexual violence problem alongside other societal developments. Change is gradually afoot, with stronger criminal provisions, the OSC, and new mechanisms for content removal and account restriction. However, so much remains unresolved. From normalisation and gendered entitlement to uneven platform response and survivor burden, these developments are further amplified by technological change. I wonder whether these changes can, over time, shift responsibility towards perpetrators, restore survivors’ agency, and reshape familiar social scripts.

Many media reports and much of the public discourse have focused on the paper’s provenance, specifically the names listed in the acknowledgements section and the Singaporean academics who have “distanced” themselves from involvement with or acknowledgement in the paper. And as I read through the paper in detail, I observed three reporting gaps:

1. There are few substantive critiques: The aforementioned academics and experts rarely offer substantive assessments of the paper and its findings. Even when they do so (I spotted a small handful), the assessments are very brief and thin, which doesn’t help the lay reader understand the strengths of the paper’s arguments.

2. Reporting of the paper’s arguments is only half complete: The authors presented two research questions (and conclusions). Some of the media reports explained the first well, but most paid no attention to the second. The paper is titled “Do Social Norms Substitute for Enforcement?”, and, as it stands, the most widely reported finding isn’t the question presented in the title.

3. The data-related questions are fascinating yet not fully engaged with: The authors use and combine multiple data sources, and it is the database from a proprietary source “which contains demographic details and residence addresses for a large sample of Singaporean residents as of 2013” (p. 18) that interests me the most.

(I empathise with the challenge of dissecting empirical analyses in academic publications. I have some expertise in quantitative research but used an LLM to work through the framework and model detailed across pages 7 to 14 and the corresponding appendices. Even so, the paper warrants further attention and scrutiny.)

Against this background, having spent a little more time with the paper, I hope to do three sequential things. First, briefly explain how the authors inferred informed purchasing among civil servants and establish why I find the arguments convincing. Second, in a similarly accessible manner, explain the second research question and conclusion that social norms appear not sufficiently self-sustaining to substitute for continued formal enforcement. And third, with the Public Service Division (PSD) saying that it is reviewing the paper, I explore some questions related to the evidence, data, and data source as well as competing explanations.

From unusual purchases to informed purchasing

Here’s the widely reported finding in a nutshell. The authors found that people they identified as civil servants purchased private homes near future MRT stations at unusually high rates one to two years before those station locations were announced to the public. Mid-level officers and those connected to rail-planning agencies were more likely to make these unusual purchases, which appears consistent with informed purchasing using non-public information.

The data suggest informed purchasing among civil servants. However, did the civil servants use non-public information to buy those private homes near MRT stations, or were those civil servants already savvier or more discerning – or were they more likely to be high-income earners, which would allow them to afford more expensive private homes – compared to non-civil servants? Note that the authors do not use direct individual income for the matching (more on this later). Therefore, the authors had to construct matched non-civil-servant comparisons. As a stylised illustration using averages from Table 1:

So now, after this matched comparison, we find that the unusual home purchases happened among the civil-servant group during a time advance information would have been useful (and potentially profitable). In other words, there is a correlation between being a civil servant and buying private homes near future MRT stations before the station locations were made public. Yet, as we’ve all heard, correlation does not imply causation.

What else could explain the pattern?

This is, in my opinion, the paper’s strongest section, where the authors test alternative explanations to strengthen their interpretation (i.e., What else could explain this pattern or correlation?) This is also why I find the overall argument about informed purchasing fairly convincing. CNA did a good job stating some of these checks, and I’ve synthesised them here:

1. Maybe the results arose simply by chance? The authors randomly relabelled who is deemed a civil servant and also randomised the MRT locations and announcement timings. “Evidence of informed trading by civil servants is unlikely to hold by chance” (p. 23).

2. Perhaps civil servants are more likely than others to want to live near future MRT stations? The authors repeated their analyses with HDB resale flats and found no comparable pre-announcement civil-servant pattern.

3. Are civil servants just unusually good at reading and responding to public information? The authors compared real estate agents and corporate board members, and, similarly, the repeated analyses did not reproduce the pattern observed among civil servants.

Finally, if one buys the argument that civil servants having access to privileged information influenced their purchasing decisions, then you’d logically expect officials of agencies involved in the planning of rail line systems to exhibit the strongest pattern of informed purchasing. And that is precisely what the authors found.

Share

(I spotted some cases of academics and experts engaging substantively with the paper and its findings, though I’m not sold on their scepticism. One said that the potential “gain” from private homes, which are usually farther from MRT stations vis-à-vis public housing, is “much smaller.” Another argued that future MRT stations may be reasonably inferred from construction works in an area. Both are plausible alternatives, though they must also explain the distinctive timing and the unusual patterns observed among civil servants. The paper does address some but not necessarily all of that concern.)

The authors present evidence for potential economic motivation too. They find that the civil servants do not pay a significant premium on their purchases, and the purchased properties – held for shorter periods, which is also consistent with an investment motive – are later sold with above-benchmark returns. Moreover, connected “relatives,” defined as co-registrants or co-occupants sharing the same registered address, also showed similar patterns of informed private-home purchases.

All the evidence, taken together, renders the current interpretation more persuasive than the initial correlation alone.

“Do social norms substitute for enforcement?”

At the same time, the most widely reported finding – which I detailed in the preceding section – isn’t actually the question presented in the title. For decades, incorruptibility has been Singapore’s calling card and cases of public corruption are perceived as extremely rare. The combination of strong anti-corruption institutions and strong social expectations of public-sector integrity has created pro-integrity social norms, and the authors wonder whether strong norms can substitute for costly formal enforcement to deter misconduct by public officials.

This is the second research question, and the paper concludes that social norms do not necessarily substitute for enforcement. The authors argue that formal enforcement makes misconduct costly for public officials themselves and helps preserve the value of being regarded as honest. Complicating this is the government’s inability to observe every act of misconduct and assess whether the norms remain durable (and if enforcement is needed).

How did the authors reach that conclusion? They first described 2011 to 2012 as a period of enforcement tightening, including – in their telling – the Securities and Futures (Amendment) Act 2012, revisions to the Official Secrets Act, and several high-profile corruption cases (p. 35). Then, remember those patterns of informed purchasing among civil servants and their co-registrants or co-occupants? They were evident in the earlier pre-2011 period, but no longer so in the later period, for both the civil servants themselves and their “relatives.”

Before concluding that the enforcement tightening deterred the informed purchasing misconduct, the authors further drew from public perception survey data indicating increased perceived/witnessed corruption, rising distrust in public institutions, and lower willingness to say that dishonest conduct is “never justifiable.” All these indicators later recovered. The authors do not engage with the 2011 general election, yet GE2011 – when the ruling party’s share of the popular vote fell to its lowest since independence – featured powerful signals that many Singaporeans were unhappy. Then Prime Minister Lee Hsien Loong had acknowledged concerns over policy substance and implementation, including housing.

For the authors, with the deterioration in the survey-based integrity measures, the patterns of informed purchasing suggest that some underlying behaviour also looked worse. In this vein, the authors argue that the evidence is more consistent with the government learning that social norms were less durable than previously believed. Critically, the informed-purchasing patterns declined when stronger enforcement was enacted.

What I’d still like to know

As convinced as I may be of the paper’s arguments, the role of the media – alongside academics and experts – is to rigorously review the data and methodology, before deciding if those arguments and claims are supported. If a non-domain expert like me can make sense of all these, then we, collectively, should be able to answer and pose more questions.

Take the data and data source as examples. In particular, among the multiple datasets, that unnamed proprietary source from 2013 – with demographics and residential addresses – is doing a lot of the heavy lifting (The Straits Times has asked the authors for details.) There are so many outstanding questions. What exactly is the source and where did it come from? How were the names linked and what validation was performed? Given the inferred granularity of the data, what data-access, research-ethics, and privacy arrangements were applied?

Recall too, when I described the paper’s matching procedure with the hypothetical Peter and Sam pair? The authors matched the home buyers on age, gender, marital status, ethnicity, and residential location. There is a potential methodological limitation here, because you can only match on data that you have. Other socio-demographic indicators of interest – think highest education, income and wealth, number of children or dependents, or anything that may make it more likely for someone to disproportionately buy near future MRT stations before a public announcement (e.g., someone more highly educated could better infer construction patterns) – were not reported as part of the matching specification.

Subscribe now

Any such omitted indicator or factor would also need to explain the specific pre-announcement and near-future MRT station pattern of the home purchases.

I have two other data-related questions. First, what did the authors mean by “rail-planning agencies?” Precisely which agencies or departments? Which agencies were classified this way, and using what rule? Second, while the authors explained how they identified the real estate agents, what was the source or database for the corporate board members?

Overall, I also mentioned how the broader political environment in the context of GE2011 may have increased the salience of integrity-related concerns, and public officers themselves may have responded to the increased scrutiny. As another possible explanation, in 2011, the government introduced measures to address concerns over rising home prices and property speculation (e.g., extending the Seller’s Stamp Duty holding period, increasing rates, lowering the Loan-To-Value limit). Later in the year, the Additional Buyer’s Stamp Duty was introduced. Could these changes have accounted for the earlier/later difference, and could they have explained the difference between the civil servants and matched controls?

And somewhat cheekily, my reading of the paper leaves me with the impression that some people in government – including civil servants and policymakers – were receiving signals of a deteriorating integrity environment. Stronger enforcement then followed. Was anyone aware of this specific property-purchase pattern? Were there internal audits or cases? What actually happened?

It’s clear that there’s still plenty – data provenance and linkage, how planning agencies are defined, the exact enforcement chronology, among others – that I’d still like to know. At the same time, the findings of the first research question are persuasive to me, and the second research question feels intellectually important. Given that PSD can examine administrative information that we, as outsiders, cannot, I believe experts should continue playing an important role. Academics ought to scrutinise the data and methods, while the media can report substantive findings rather than provenance alone.

(Disclaimer: All opinions expressed are solely my own and do not represent the views of the Singapore-based social service and non-profit organisations with which I work or consult. The programme examples are hypothetical.)

In my day job as an assistant professor, my working assumption – regardless of how my students say they feel about artificial intelligence (AI) and institutional policies about large language models (LLMs) – is that generative AI will be used in any take-home assignment. Only with supervised in-person tests, for instance, can I guarantee that no AI tools were used during the assessment.

I am pretty tech- and AI-savvy, but my students will always be half a step ahead. I initially tried hidden prompts, injecting “invisible” text meant to fool students. I designed supposedly AI-proof assignments, with students searching for and summarising journal articles, because LLMs still hallucinate and sometimes cannot directly retrieve paywalled articles. And I tracked revision histories of documents while using the AI-checker Pangram, before realising, ultimately, that I was losing the plot.

Today, when I read and review philanthropic grant proposals, it all feels like déjà vu. My working assumption is the same: any submission involves generative AI. Need to describe a social problem and size the population? ChatGPT’s “Deep Research” tool can generate a one-pager complete with APA citations. We need a logic model and/or theory of change for the proposed programme? Claude can populate the columns and even “beautify” the diagram. Then, we need to respond to a barrage of clarification questions? Let’s have Gemini respond for us.

There are two immediate pushbacks to my AI critique, and I am very sympathetic to both. First, so many people in Singapore are using AI, and the government is incentivising everyone to do so even more. Grant-makers themselves are using AI (e.g., many U.S. foundations and non-profits have adopted AI). If the entire system is set up to be AI-friendly, why fault organisations as grant applicants? Second, and relatedly, grant writing is a skill and grant proposals are often onerous. Overworked practitioners in these organisations don’t necessarily have the training, time, or bandwidth to perfect a grant proposal, especially when there is no guarantee that it will be funded.

If so, then isn’t AI usage less a problem than a playing-field-leveller when outstanding organisations and practitioners are no longer handicapped by their grant-writing inexperience or weaknesses? However, my critique is more pointed. Instead of trying to fruitlessly police whether, how, or when AI is used, perhaps we should accept the growing obsolescence of the grant proposal – in its current format – and embrace new ways of assessing organisations and their programmes.

And what if, alongside grant-making per se, the grant proposal and grant assessment process are also opportunities for meaningful programme development and evaluation?

“When LLMs write the proposal, and LLMs read it too”

So that we’re on the same page, social service agencies and non-profit organisations prepare philanthropic grant proposals to secure funding. In Singapore, proposals are commonly submitted to government agencies, statutory boards, companies, intermediaries, and/or private donors and foundations. Here, I use the term “grant proposal” broadly across both private philanthropy and public-sector social-service settings.

The specifics and templates of each proposal vary, though they commonly include the following components:

1. Social problem and population: You describe a problem (e.g., social isolation) for a population or subgroup (e.g., older adults in Ang Mo Kio).

2. Programme and activities: You explain what your organisation plans to do or has been doing (e.g., a weekly befriending service), and how you will implement the activities (e.g., recruit student volunteers trained by social workers).

3. Logic model and theory of change: You present your inputs/resources, programmatic activities, outputs, outcomes, and impact, explaining how and why the activities are expected to produce the outcomes (e.g., increased social connectedness and reduced loneliness among the older adults).

4. Monitoring, evaluation, and learning (MEL): You specify how the organisation will track different measures before, during, and after a service user’s participation, and how the findings may be used to improve the programme.

5. Other documents: (a) organisational profile; (b) budget; (c) financial statements.

When the components are laid out like that, even to the untrained eye, it is quickly and plainly obvious that, with text-based AI tools, the grant proposal has become a much less useful signal for assessing the potential effectiveness of a programme and an organisation’s ability to implement it. Previously, a coherent and polished grant proposal indicated organisational effort and commitment, suggesting that the organisation had thought through and responded to questions germane to a programme’s success. Now, if AI can provide ostensibly high-quality responses without the same organisational deliberation, it is no longer clear – at least to me – what a grant proposal demonstrates.

(Don’t just take my word for it. Copy and paste the above components, write a simple prompt – maybe even uploading a blank application form – and have an LLM draft a complete grant proposal for you.)

Share

What then is the “supervised in-person tests” equivalent for grant assessments? These alternatives can complement each other (and the AI-assisted grant proposal), yet they have their drawbacks. We can organise calls or in-person interviews, but some applicants may be more comfortable putting their thoughts on paper or feel anxious during face-to-face and ostensibly high-stakes meetings. Site visits – either scheduled or unannounced – can help ascertain that programmes are running as intended with appropriate participant numbers, but what happens to pilot initiatives that are still being developed?

We could amend – and, in some cases, have amended – grant application forms, with questions asking for more specific indicators. When asked whether an organisation has been tracking programmatic outputs and outcomes, for example, an LLM can churn out a verbose response detailing hypothetical constructs and measures. However, it may be more productive to ask applicants to share an appropriately anonymised screenshot of participant-management system, demonstrating clearer evidence of MEL commitment.

Are we more data-driven or trust-based?

How have we come to rely on the grant proposal? We seem to agree that grant-making that is too relational – giving to familiars, making private pleas, donating within closed circles – doesn’t seem aligned with an increasingly evidence-informed or evidence-based movement, in terms of giving to causes doing the most good. Therefore, grant proposals are a way to obtain data and information and make more “objective” decisions.

In April 2025, on her podcast On with Kara Swisher, American journalist Kara Swisher interviewed businesswoman and philanthropist Melinda French Gates (a transcript is available). During the interview, Swisher asked Gates to describe the differences between a data-driven approach to philanthropy (championed at the Gates Foundation) and a trust-based one (popularised by novelist and philanthropist MacKenzie Scott, who has donated tens of billions of dollars in unrestricted funds with limited application and reporting burdens).

Gates described the trust-based approach:

And the data-driven one:

The convenient answer, when presented with two approaches or models, is to acknowledge that they are not mutually exclusive, that it is not an either-or binary. Why box us in, when we are positioned somewhere in the middle (this was also Gates’s answer in terms of her own philanthropic style)?

That answer feels unsatisfactory to me, because my follow-up would be: then, which approach are we more aligned with? Because whether we are more trust-based or data-driven will likely shape how we design the grant proposal and corresponding grant assessment process, and whether we see AI proliferation as an opportunity or a threat. Very generally and collectively, at least in Singapore, we appear much more comfortable with data-driven approaches – and with good reason, given concerns such as accountability and governance – which renders us more likely to still cling to the grant proposal and all its trappings (e.g., applications, reporting, progress updates).

So… what next?

And perhaps this commentary feels like a bait-and-switch? I started by problematising AI usage among grant applicants and grant-makers – LLMs writing grant proposals, and LLMs reading them – and now… we’re assessing philanthropic approaches?

Because I think short-term AI fixes exist. In addition to the aforementioned complements or alternatives – the calls, in-person interviews, site visits, and more specific indicators in amended forms – organisations and funders (and even the sector) can set guidelines and/or thresholds for acceptable LLM usage. For example, at least 70 to 80 per cent of proposal text should be human-generated (this proportion is completely arbitrary right now). We may have applicants sign honesty pledges or specify that LLMs should only be used for brainstorming at the start and copyediting at the end. If applicable, proposals could detail how LLMs were used and include the prompts.

I must also add that there will be a perceived double standard, if funders impose restrictions on AI usage among non-profits, but themselves use LLM for grant assessment. If we accept the grant proposal as critical, both the funder and agency must invest their own time and effort in the process (as we do in the classroom).

And so, what if the AI onslaught presents an opportunity for grant-makers and grant applicants to tweak (or reset) a relationship historically characterised by disproportionate power dynamics? Or to inject a little more trust – with a lot more collaboration – in a predominantly funder-led and data-driven Singaporean paradigm? In this vein, the grant proposal no longer functions as a summative assessment, with grant applicants (as the “student”) doing the lion’s share of the work, and funders (as the “teacher” or “professor”) deciding whether the proposal passes muster.

Consider a combination of these options:

1. A very simple grant proposal, at the very start, that prioritises due diligence and some baseline expectations. After all, even Scott’s high-trust, low-control giving still involves assessment. Her quiet-research process considers finances, track records, outcomes, and leadership, while her 2024 Open Call used an application.

2. After funders establish that these are organisations they can work with, the relationship becomes less subcontractor-like and more collaborative, with the “grant proposal” – as we currently know it – jointly produced and refined by both parties. Writing is hard! Crafting a coherent programme narrative takes time and effort. Conversely, reviewing and offering comments is a comparatively easier lift than developing the initial programme narrative. In other words, during this initial phase of conceptualising or developing a programme, funders could work with the non-profits, rather than delegate most of the proposal development process. We could work through assumptions, develop the intervention, or design a feasible evaluation.

For funders, this would mean require a commitment of staff time and relevant expertise.

3. Progressively, the grant proposal functions less like a one-sided accountability mechanism and more like a joint agreement on what funders and non-profits want to implement together and how implementation will be collectively tracked and measured, such that grant-making functions more like a collaboration.

4. Such cooperative approaches, wherein funders have more skin in the game (above and beyond their financial capital), may increase the appetite for experimentation and “riskier” bets, when we are less fixated on identifying and funding that “perfect” grant proposal.

Then, deliberations over whether, or to what extent, AI is used become less critical, because the grant proposal is less likely to be perceived as an assessment to pass. Instead, the focus turns to developing and evaluating the best version of a well-intentioned programme – implemented by a well-resourced organisation – for the benefit of service users. Establishing programme understanding is a more meaningful endeavour than checking authorship!

Subscribe now

Today’s version of the philanthropic grant proposal, given the pace of development of text-based AI tools and LLMs, is rapidly losing its value as a signal for high-quality programmes and organisations. Funders may respond by policing AI usage in the short term. Unfortunately – as I’m experiencing in the classroom – we’ll always be half a step behind, and back-and-forth exchanges to check if LLMs were used feel time-consuming and counterproductive, with the potential to sow mistrust. Furthermore, these processes may privilege those with access to more sophisticated AI tools, who could be better able to navigate checks and detectors and other barriers to entry.

I also recognise that a wholesale turn to trust-based approaches isn’t a panacea, especially within a Singaporean context with a persistent fixation on KPIs and without sufficiently developed muscles for programme development. I do think that changes related to what goes into a grant proposal, how a grant proposal is drafted, iterated, and assessed, and, ultimately, how it is used are timely and much needed. Put otherwise: I want funders collaborating in programme development rather than mainly assessing the resulting proposal.

It was also the Father’s Day weekend, including in Singapore. As we celebrated fathers’ contributions and sacrifices, it felt timely – even as many are doing more in Singaporean homes – to ask what would allow fathers to be more present.

Most recently, sociologist Teo You Yenn’s multi-week bestseller about parenthood and doing family included a chapter on the uneven burdens and consequences of work and care for men and women (“Unease: Life in Singapore Families,” Chapter 4). Her interviewees reflected “that men today do more housework and caregiving than their own fathers did.” However:

“The big picture will be obvious to anyone living in Singapore today: women take on more housework and caregiving than men do, even when they also do wage work; men carry greater responsibilities to be wage earners, even when their wives work.”

Yet, even against this unequal backdrop, with household duties disproportionately borne by mothers and fathers facing provider expectations and workplace norms that make caregiving harder, Singaporeans overwhelmingly – and I mean, overwhelmingly – believe that the status quo should not hold. Among married respondents, 99 per cent agreed that fathers and mothers are equally important caregivers. And when it came to the home, 95 per cent of married respondents and 98 per cent of single respondents agreed that both parents should share responsibilities at home equally.

We believe care should be shared. At home, care remains gendered.

The data on perceived parental responsibilities come from the 2021 Marriage and Parenthood (M&P) Survey, involving approximately 5,900 respondents (2,848 single or never-married and 3,017 married Singapore residents). Overall, attitudes in Singapore have become more egalitarian. More respondents now reject the idea that mothers should be the default full-time caregivers and more say that household and caregiving responsibilities at home should be shared.

Based on what respondents say, from 2016 to 2021, those agreeing that “Ideally, the mother should take care of her children full-time,” fell from 56 to 40 per cent among married respondents, and from 40 to 24 per cent among single respondents.

Singaporeans believe and say care should be shared, yet actual practice and action still lag behind these beliefs and attitudes. In the same section of the 2021 M&P Survey, it is stated that “married women continued to do more at home compared to their husbands, primarily in the area of child-related care.” Additionally (emphases mine):

The unequal sharing of care features in both childcare and household work. In 2016, mothers reported spending 2.6 hours on domestic chores on a normal weekday, compared with 1.5 hours for fathers.

Share

SG-LEADS, the Singapore Longitudinal Early Development Study, is a nationally representative longitudinal panel study, which means it has tracked and repeatedly collected data from a sample of households with Singaporean children since 2018. In its June 2022 update, using data from the Wave 2 survey conducted in 2021 (involving 4,352 Singaporean children and their primary caregivers), SG-LEADS presented findings about the time spent engaged directly with parents on weekdays and weekends. Relatedly, the Wave 2 report showed that 95.7 per cent of primary caregivers were the children’s biological mothers.

Unsurprisingly, when comparing the time children spent with their mothers and fathers, across the categories shown – caregiving, play, social activities, achievement-oriented activities, and travel – children spent more direct time with mothers than fathers, though the size of the gap varied by activity. In the categories shown, fathers do not exceed mothers in direct time spent with children.

Everyday care remains gendered, even as attitudes have moved strongly toward shared caregiving and many fathers want to be present. Today's fathers are more present, and likely more engaged than their own fathers, but mothers still carry more care work and responsibility. Moreover, the conditions that would allow fathers and mothers to share household and caregiving duties more equally are also uneven.

Paternity leave shows what policy can – and cannot – do

Given these disparities, it is tempting to fault individual fathers for not stepping up. Individual accountability still matters, especially where fathers refuse to match their partners’ time and effort. More broadly though, fathers in Singapore – like mothers and parents in general – are constantly navigating constraints. If we want fathers, as parents, to spend more time at home and with their children, then financial pressures, work schedules, work-life balance, affordable care, support systems, and social expectations matter too.

In the 2021 M&P Survey, when married respondents were asked about why they decided not to have any more children, the top six reasons – based on those who selected the reason among their top three – were financial cost (64 per cent), having enough children (50 per cent), the stress of raising children in Singapore (49 per cent), difficulty managing work and family demands (34 per cent), the lack of good caregiving arrangements (17 per cent), and difficulty conceiving (14 per cent).

Towards the end of the aforementioned chapter in her book, Dr. Teo expounds on the disconnect between Singapore’s ostensibly “family-friendly” policies and the realities of (gendered) inequalities, and how this contributes to housework and care labour being treated as low-status, feminine labour:

“Men have not stepped up in the domestic sphere in the ways women have stepped up in the earning sphere. Women, including those who can afford paid care, remain responsible for resolving care needs, and this impedes their workplace advancements. Importantly, the issue of reconciling work and care, although widely shared and a consequence of significant social transformations, is cast as a ‘private’ problem to be sorted out by individual families and usually by ‘the woman.’”

For me, the data around government-paid paternity leave are illustrative as a test case. The broader literature is clear that father involvement matters for child development and family well-being. Singapore research using SG-LEADS data makes the case more concrete. Children whose fathers took at least two weeks of paternity leave had better problem-solving and word-recognition skills, as well as fewer behavioural problems. Fathers felt closer to their children, mothers reported lower stress and higher marital satisfaction, and family relationships improved. Sure, paternity leave alone may not necessarily encourage couples to have more children, but it confers enough benefits to fathers and their families for it to be a no-brainer (and its value does not just rest on fertility).

Paternity leave is not about fathers “helping” mothers, but fathers being recognized, expected, and supported as parents from the beginning. Recorded take-up of Government-Paid Paternity Leave rose from 47 per cent for children born in 2016 to 61 per cent for children born in 2024. There is an important measurement caveat, because the proportion excludes cases in which employers did not submit claims

Today, eligible working fathers of Singaporean children born on or after April 2025 are entitled to four weeks of government-paid paternity leave. Eligible working parents are also entitled to six weeks of shared parental leave for children born from April 2025 to March 2026, rising to 10 weeks for children born from April 2026.

Which fathers get to be involved?

The easy conclusion is that fathers do not care enough, or that they are happy for their wives to shoulder most of the childrearing duties in the first months of their children’s lives. The more complicated and less satisfying interpretation is probably that we do not know enough. Are fathers deterred by workplace perceptions, if they are seen as less committed, or if their colleagues think they are being imposed upon? Do they anticipate any formal or informal penalties in relation to their present and future career prospects? Would take-up improve if more of their peers took leave, leading to broader shifts in culture? How do mothers, grandparents, domestic workers, and employers shape fathers’ choices?

The available evidence points to class and occupational stratification, especially in leave-taking. Blue-collar workers appear less likely to apply for paternity leave because of job concerns. Similar concerns are likely shared by lower-wage, shift-based, contract, or operational workers, who may face higher risks when taking leave or asking for flexibility. For instance, “fathers with machine operator, assembler and cleaner jobs were less likely to take paternity leave,” and such fathers may further experience long work hours and a male-dominated work culture.

A Father’s Day wish list for fatherhood research and practice

For a social service sector that wants to bridge the gap between belief and practice, the first step is to identify what we do not yet know. Globally, similar to the trends in Singapore, more are recognising the importance of fatherhood and fatherhood research:

“According to some estimates, the average time dads spend caring for their kids each day has quadrupled over the past 50 years. Their attitudes about parenting are also changing. Today, men are about as likely as mothers to say parenting is a key source of meaning and a central priority in their lives. Roughly 85% of fathers identify parenthood as one of the most important aspects of their identity.”

(This is not just an abstract interest for me. As a postdoctoral fellow at the University of Denver, I spent two years supporting the evaluation of the “Nurturing Dads and Partners” programme, which integrated peer and co-parenting support to build fathers’ skills and family protective factors. I remain deeply involved in fatherhood-related research.)

For a start, here is a list of father- and fatherhood-related questions worth exploring:

1. Paternity leave: What predicts fathers’ actual use of paternity leave, especially among lower-wage and shift workers? Who are the fathers not using their leave? What prevents them from doing so? For instance, in terms of cultural scripts, how do their employers and supervisors as well as peers respond when fathers take leave?

2. Financial considerations: Given that finances and work-family demands are central to parenthood decisions, which anxieties around the cost of living, income security, and workplace predictability shape fathers’ caregiving decisions?

3. Social work practice: How do family coaches, social workers, and teachers and other school staff engage fathers compared with mothers? In ComLink+ and other family-support programs or settings, which parent is (first) contacted, who attends meetings, and how are fathers invited in without displacing or burdening mothers further?

4. Father-specific programmes: What types of father-specific interventions work in Singapore, and for whom? How effective are one-on-one or group-based fatherhood programmes? For fathers who may have been less involved, absent, or navigating divorce and separation, how might co-parenting courses be useful?

5. Alloparenting, or “It takes a village”: What if fathers and mothers are not expected to do everything on their own? How might alloparents – other caregivers who share parental care, such as grandparents, relatives, neighbours, and paid caregivers – share the responsibilities? In Singapore, any serious discussion of “It takes a village” also has to reckon with our reliance on migrant domestic workers inside the household.

(Dr. Teo examines in detail the assumptions upon which Singapore’s social welfare policies are built: self-reliance, familial support, and “Many Helping Hands”.)

Subscribe now

Discussions of fatherhood are inevitably intertwined with discourse about masculinity and expectations of men in contemporary society (e.g., “Of Boys and Men”). In Singapore, since 2009, “Dads for Life” describes itself as a national movement “to inspire, mobilise, and involve fathers.” More recently, a coalition of 12 men’s groups launched “MenToo,” a movement that “wants Singapore men to speak more honestly about mental health, relationships and the pressures they carry” and is also complemented by a research collaboration. (I agree with Jom that the movement is worth taking seriously, even if the name risks being read as a negation of #MeToo.)

The Japanese World Cup image states, plainly: “Please do it at home.” In Singapore, for me, that means asking whether work, leave, income security, family support, and social service practice make care shareable among parents. Dr. Teo’s reminder is useful here, that care should not be treated as a private problem for women to solve.

Rather than ranking fathers and mothers against each other, we want to make caregiving more shareable, so that mothers are not left carrying the invisible load or paying the “motherhood penalty,” fathers are not trapped only in breadwinner or provider roles, and Singaporean children receive care from adults who have the time and support to be present.

Having appreciated the general importance of evaluation, now, many of us in the sector are attending the same introductory courses, using the same logic model templates, saying outputs alone are inadequate and that outcome measures are more important, crowing about, one day, calculating deltas and counterfactuals and social return on investment, yet ultimately falling back on anecdotal data when all else fails.

That is, we’re doing the same things over and over again, often without corresponding evaluation of our evaluation efforts, and expecting different results. For instance, our learning of the tools and vocabulary has not translated into significantly better programme designs. Additionally, the Sector Evaluation Framework (“SEF”) by the National Council of Social Service is necessary, yet insufficient. It is most useful as a shared introductory scaffold. The framework provides the sector with a common language, though a common language can become constraining when programmes, communities, and theories of change are increasingly diverse. The framework is also less helpful when programmes and interventions don’t fit neat, standardised indicators.

Unlike Sisyphus, we’re not bound to the same routines and frameworks. The rut has three parts: (a) a weak evaluation infrastructure; (b) over-reliance on logic models; and (c) a still-underdeveloped evidence ecosystem. For a start, Singapore’s social service sector needs a well-staffed evaluation infrastructure, not more individual evaluation literacy. It also needs programme development and evaluation approaches that don’t fixate on the logic model per se. Over time, even as organisations improve their capacities to collect and analyse data, external evaluators, university researchers, and funders must help shape such processes and related incentives.

Individual evaluation literacy is not evaluation infrastructure

Because we’ve been too fixated on individuals and their evaluation literacy, organisations are not necessarily set up for success. At the moment, among smaller social service agencies, evaluation work is often a one-person show, led by someone who is double-hatting or reliant on transient volunteers or interns, and ultimately unsustainable when these over-worked practitioners – no matter how well-trained they may be – leave the organisation or sector when they’re exhausted or burnt out.

For larger agencies, leadership and board alignment appear to be decisive. Larger agencies whose leaders who see evaluation’s strategic value – for example, in surveying service users to ascertain programme demand (i.e., needs analysis and formative evaluation) or being more deliberate about scaling evidence-informed programmes (i.e., outcome evaluation) – are more likely to view evaluation as intertwined with day-to-day operations. Besides the immediate implication that evaluation literacy efforts ought to target EDs/CEOs and board members, organisations can only rigorously engage in evaluation with adequate manpower. Otherwise, evaluation is construed as an imposition or sai kang.

Organisations need to start with an evaluation headcount: Within organisations, evaluation and practice research – which aims to generate research evidence to inform practice by involving practitioners, researchers, and service users – will only be prioritised when there is a job description staffed by someone. If there’s no headcount, or even a full-time staff member with partial responsibility for evaluation (and this cannot simply be the ED/CEO), then any capacity-building effort will be short-lived.

Increasingly, organisations are not unaware. In fact, they are already sold on the merits of evaluation. However, the problem persists with a combination of factors: pressures to keep administrative overheads low and operations lean, funder reluctance to support manpower expenditures, preferences to channel resources more directly to service users (or the tendency to separate programme costs from learning and training costs), a lack of evaluation expertise in the job market… More on these later.

Subscribe now

The sector must offer ongoing training and development pathways beyond introductory literacy: Even when there is an evaluation headcount, that person (or half a person) is frequently the only one in the organisation – besides the leadership team – who is engaged in evaluation and practice research. For everyone else, the demands of evaluation-related data collection and analysis are inconveniences. (Which is understandable, given social workers’ expansive and growing workloads.)

For research-minded practitioners who want to go deeper, the introductory courses are too elementary and generic, and they are not sufficiently customised. Beyond structured courses, sector-specific communities of practice, peer learning, and methodological depth may be lacking.

Practically, there are very few intermediate or advanced training courses for these practitioners. Many alternative courses are offered in universities, and even though they are subsidised, they remain very research-heavy. Ultimately, the sector needs methodological depth that is practice-facing, not only academic.

And how might these courses be made more practical? Currently in the sector, qualitative evidence might be dismissed as less rigorous or less “objective.” However, the challenge is that organisations and practitioners don’t have the ability and expertise to use qualitative methodologies to generate practice-research insights. Put otherwise, the problem is not qualitative evidence. The problem is unstructured anecdotes presented uncritically as qualitative evidence, and rigorous qualitative work being dismissed as not “objective” enough. Furthermore, larger agencies and their leadership may benefit from the use of evaluation evidence for strategic decisions.

The accursed logic model has substituted programme theory

In 2011, Laurence Lien – then chief executive officer of the National Volunteer and Philanthropy Centre – characterised most Singaporean non-profits as subcontractors, delivering social services as contracted by the government. For Mr. Lien, this was a missed opportunity, because non-profits could do what the government cannot do (i.e., cater to diverse and holistic needs, address emerging issues, offer niche and customised solutions), and even do some things better (i.e., innovate and experiment, use moral authority and convening power to engage with each other and service users). He wrote:

“The brains and heart of social intervention remain with the State, while [non-profit organisations, NPOs] often simply follow the piper’s tune. NPOs trying to effect change through competing models of intervention are often viewed as threats. Hence, many NPOs lose their own sense of aspiration, and contribute insufficiently to support necessary social change.”

Today, elements of this contractor-subcontractor model still seem to persist, even as the social service sector – grounded by service users and increasingly strengths-based and community-centred – has demonstrated its potential to play a larger role as a complementary co-partner. To play a bigger role. Furthermore, many socio-economic issues in Singapore, especially those affecting the most disadvantaged and minoritised, are too complex for delivery alone. For example, social service agencies with the requisite evaluation manpower can better experiment with interventions and generate actionable insights to tackle those issues.

The logic model, without theories of change, is not enough: The subcontractor paradigm also helps explain why the sector’s evaluation potential is limited. I’ve attended enough workshops and reviewed enough content to know that the logic model – presented too often as the be-all and end-all of evaluation – is used mechanistically, without firm understanding of the underlying theory of change. Relatedly, organisations, because they are required to do so, use logic models to explain and justify existing programmes, instead of interrogating key programmatic assumptions.

Technically, in particular, there is a lot of disproportionate anxiety over crafting “SMART” outputs and outcomes – fussing over the “right” measures for constructs of interest – and much less focus on the problem and change theories or the underlying assumptions that undergird programmes.

These observations are complicated by the fact that, in the sector, we don’t have the language and ability to critique logic models. We are often satisfied by the presence of one. It helps funders feel assured and organisations feel confident, whether justifiably so or not. It also explains why logic-model development and assessment can feel so arbitrary and subjective – with one funder signing off and another disagreeing completely – and why organisations err on the side of caution by cramming their logic models with as much content as possible, to the extent of rendering them unreadable and undecipherable.

Programme development is a lost art: Another explanation for how the logic model is mechanistically used, primarily to explain and fine-tune existing programmes, is that the sector has lost the muscles to innovate, develop, and pilot new programmes. If elements of Mr. Lien’s critique still hold – that social service agencies continue to largely be subcontractors, taking on ready-made programmes that mainly require implementation – then there are few incentives for programme development.

In other words, logic models are used to justify existing programmes because agencies are positioned as implementers (or subcontractors delivering programmes), not programme or intervention designers (or partners supporting experimentation and adaptation).

Being stuck at the logic-model phase without corresponding programme theory and depth affects agencies differently. Ground-ups or smaller agencies with good proofs-of-concept or pilot programmes may not scale effectively. Or, more commonly, these organisations do not gain enough traction in the sector to obtain the requisite funding or resources. Larger agencies that maintain a portfolio of existing programmes may sustain those while riding on their track records yet struggle with producing new programmes. What perhaps remains underdeveloped is the middle space between promising pilots and institutionalised service delivery.

Singapore’s evaluation ecosystem is still underdeveloped

As I alluded to earlier, organisations operate within a nascent evaluation ecosystem where the actors exist. Even so, the incentives, resources, and outputs are not well aligned.

First, external evaluators are too expensive for most agencies. There is a small but growing pool of external evaluators in Singapore – such as Research for Impact, Soristic, the Centre for Evidence and Implementation (Disclosure: I know and have worked with these organisations and others) – and while I believe their services, given the expertise and manpower involved, are priced correctly, the prices are still too steep for the overwhelming majority of charities. Without government or philanthropic support, even larger agencies have to prioritise the few programmes to evaluate. Smaller agencies rely on pro bono rates or generalised support like Evaluation 101 workshops.

Even when agencies can afford external evaluators, as I am learning myself, they still need clearer models of collaboration. We debate who owns the questions and data, who interprets the findings, who makes and presents recommendations, and how the learning will be used.

Second, university researchers have differing incentives. University researchers are a plausible alternative to external evaluators, yet we have different incentives. More often than not, academic knowledge production is not aligned with practice learning or practice-research applications. As an academic on the tenure track, I am familiar with the expectations to publish in peer-reviewed journals (with long review and publication timelines, compounded by esoteric writing styles largely inaccessible – literally and figuratively – to the sector), present at academic conferences (with presentation content primarily designed for our academic peers, not research-minded practitioners), while also teaching, obtaining research funding, and fulfilling service obligations.

Even with the misalignment of incentives, there is a small contingent of university researchers in Singapore who seek to work with or embed themselves within agencies. Still, the lack of time and capacity remains a constraint.

Share

Third, evaluation and evidence standards continue to be funder-led. The social service sector cannot learn from programme evaluation if negative or null findings are treated as organisational failure. (They are not.) Given the government’s central role in funding and regulating the sector, agencies positioned as subcontractors may fear being penalised for failure. If positioned as co-partners, agencies have more room to learn from failure.

Agencies themselves may want to evaluate, but evaluation may be perceived as a threat to funding continuity. If so, funders should help absorb the risks. Otherwise, social service agencies default to what is safe and conservative.

Fourth, more broadly, data availability and infrastructure – including SSNet access and duplication, data-collection fatigue among practitioners, and survey fatigue among service users – should be addressed. Agencies may also face duplicated data entry, slow or difficult data access, and uncertainty over how to turn collected data into decisions.

Given their workloads and limited bandwidths, for social service practitioners, data collection competes with casework. For service users, research and evaluation in the form of data collection (e.g., completing a survey or participating in a focus group discussion) can feel like another assessment or retelling of hardships, or even extractive. Smaller agencies may collect data without analysing them or converting them into insights. For larger agencies with some data collection experience, the results or insights from the data gathered may not be immediately clear. As such, their next goal may be to strengthen data analysis and convert findings into decisions and programme adaptations.

Finally, even when useful evaluation and research are done, findings often remain trapped in funder reports, internal decks, or academic outputs. They do not always become usable sector knowledge. Relatedly, the sector also has to prioritise stronger incentives for sharing learning over agency-to-agency competition.

The way forward: From compliance to learning

Given that the sector needs a well-staffed evaluation infrastructure, not more individual evaluation literacy, then the immediate calls to action would include the funding of organisational evaluation roles, not just training (so that practitioners can develop skills and expertise over time) and the building of intermediate and advanced evaluation and practice-research pathways for research-minded practitioners (who can support their organisations and each other across the sector). Protected time for these practitioners should be encouraged.

In terms of the substantive content of Evaluation 101 workshops, we could keep the focus on logic models while expanding the emphasis to theories of change and programme development. More fundamentally, the medium- to long-term shifts require a reconceptualisation of the social service sector’s role vis-à-vis the government, away from subcontractor and toward co-partner. Without that, efforts to educate boards and funders will be limited.

Share socialservice.sg

When social service agencies are positioned as co-partners, funders and the government should then allow for greater programmatic experimentation, with expectations and tolerance for negative or null findings (e.g., learning from unsuccessful programmes, rather than finding fault). Qualitative evidence should be legitimised, particularly when it is systematically collected, analysed, and used. Additionally, as organisations scale internal capacities, the sector could pool access to external evaluation support, such as ad hoc methodologists, research mentors, and programme- or project-specific consultants. Part of this endeavour requires dissemination and the maintenance of a knowledge-sharing infrastructure, including local publications, seminars and conferences, and communities of practice.

In the end, if Singapore’s social service sector wishes to move beyond Evaluation 101, evaluation must be resourced, practised, and governed as part of programme development itself. This includes asking better questions, testing assumptions, learning from failure, adapting interventions, and sharing knowledge openly. Agencies, with government and funding support, must build internal capacity. Funders and government can absorb more risk. Researchers and evaluators ought to work closer to practice on the ground. And the sector as a whole should interrogate and circulate new knowledge, besides merely reporting it. Otherwise, we improve our evaluation language and vocabulary without improving our practice and capacity to learn.

In her 2024 book, “The Last Human Job,” American sociologist Allison Pugh developed the concept of “connective labour,” or work that people do for each other, forging an emotional understanding and creating valuable outcomes in the process. The connective labour of social workers, for example, relies on empathy for service users, the spontaneity of human contact between social workers and service users, and their mutual recognition of each other’s humanity. Put simply, connective labour is connective because individuals see (and help) another, and, in turn, are seen (and helped).

(I first encountered Pugh’s work through SMU sociologist Jacqueline Ho, who referenced it during a panel on Singapore’s 2026 budget.)

Even before Singapore went all in on AI earlier this year, Minister for Social and Family Development Masagos Zulkifli, in a 2025 speech, flagged the impact of technology and AI on society (e.g., social isolation and job displacement), and their corresponding impact on the social service sector (e.g., positives such as improved service coordination, and negatives like data leaks and misuse). And when the minister emphasised that social work remains centred on the well-being of Singaporeans, familiar as that language may be, it reinforces the value of connective labour and the importance of social service professionals (emphases mine):

“Ultimately, we are a sector built on heart and soul. Our use of data, technology and AI can allow us to provide more personalised and targeted care. But, the human touch and empathy with which we deliver our services can and must be evident even as we leverage technology for service delivery.”

Since the minister’s speech last year, the growing supply and quality of AI tools, including LLMs (large language models), have been followed by rising individual demand and organisational adoption. Beyond general-purpose LLMs like ChatGPT and Claude, some social workers are also using Singapore-specific digital tools like Scribe, which records and summarises conversations into structured notes), and Care360, a patient management system for medical social workers. On the horizon, SSNet – a digital platform designed to collate and store data and cases of service users – will be progressively replaced by CaseCentral, a new AI-powered system for case management.

Amid such AI advancement – while staying cognisant of general AI and LLM critiques – interrogating a range of AI threats more specific to social workers and Singapore’s social service sector will help us better anticipate and address challenges. More specifically, I see nine threats, grouped into three categories: (1) AI changing service users and their helping relationships with social workers; (2) AI changing social workers themselves and their day-to-day work within organisations; and (3) AI changing the sector’s relationships with funders, donors, and the state.

The nine threats are not predictions or prognostications that all will materialise. Rather, because AI adoption is already underway, I believe the sector ought to name them early. And as the speed and rate of change accelerate, I hope to continue documenting AI’s impact as well as the evolving responses of social workers and social service professionals engaged in connective labour.

When AI changes helping relationships

AI will alter the helping relationships that social workers have with their service users, because AI is already changing how some people seek advice, support, and reassurance. At the moment, even with limited national data, younger and female respondents in Singapore appear more likely to seek emotional support or mental health assistance from AI chatbots. Very young Singaporeans are also using AI tools for gaming and studying.

When chatbots become therapists and counsellors

Service users’ growing reliance on AI chatbots for emotional support, personal advice, and crisis guidance may alter their social relationships with family members, friends, and social workers. As virtual therapists and counsellors, chatbots could alter these helping relationships in two related ways. First, they reduce friction, even though such friction – like being gently challenged or hearing disagreement – is productive and necessary. Second, the leading LLMs are found to be highly sycophantic (i.e., flattering, people-pleasing, and affirming), including when users seek advice about interpersonal conflicts and dilemmas.

Such design choices may be reinforced by engagement incentives. Users tend to trust and prefer sycophantic responses, even when those responses distort user judgement. As a recent research study found, these mechanisms decrease pro-social intentions and promote dependence:

“In our human experiments, even a single interaction with sycophantic AI reduced participants’ willingness to take responsibility and repair interpersonal conflicts, while increasing their own conviction that they were right. Yet despite distorting judgment, sycophantic models were trusted and preferred.”

Over time, even as social workers develop improved AI literacy to better understand AI chatbots and the chatbots’ long-term impact on interpersonal connections (and how they are used for personal advice and support), they may find service users coming to them less frequently (to avoid the friction from therapeutic advice- and help-seeking), or service users having less trust in them (because of the chatbots’ persistent sycophancy).

In the most extreme situations, AI-associated suicide risk, suicide ideation, and psychological disturbances linked to intensive chatbot use are credible concerns, especially for vulnerable and younger users. This is not hyperbolic. In the US, for instance, there was a wrongful death lawsuit against AI company Character.AI, following the death of a 14-year-old boy, who died by suicide after having virtual conversations with multiple chatbots and forming an intense emotional attachment with a chatbot involving romance and sex. ChatGPT has also been sued after the death of a 16-year-old boy, whose parents alleged that an especially sycophantic version of the chatbot discouraged their son from seeking mental health help and offered help to write a suicide note.

In Singapore, the health ministry has emphasised the practical difficulties associated with tracking chatbots, and that these tools are not appropriate replacements for qualified mental-health professionals.

Subscribe now

When children and youth offload thinking too early

AI’s spread in Singapore’s higher education sector seems inevitable, and while I am glad that the education ministry is taking it more slowly for younger students, I still have my reservations, particularly as a university professor trying to discern between “good” and “bad” AI education use.

Outside school, among the youngest AI users, I am most concerned that young service users may not have the patience or ability to engage in therapy, programmes, and social work services. While I accept that AI can help learning when used as a personalised tutor, it may reduce incentives for Singaporean children and youth to think and struggle productively, and it is most risky when cognitive and socio-emotional growth are compromised. Over time, AI may alter the developmental capacities that young service users need to engage in helping relationships.

A recent piece in “The New Yorker” presented three relevant cognitive and socio-emotional risks. First, LLMs may encourage cognitive offloading, before children and youth have done cognitive onloading and developed metacognitive abilities. Second, as aforementioned, AI chatbots warping their selfhood and relationships through mimicked emotional intimacy and sycophantic tendencies. And third, AI “confuses ends and means” by prioritising efficient shortcuts over messier developmental experiences.

In Singapore, for youth workers and school social workers, I am concerned that young people may not retain the patience and capacity needed for help-seeking itself.

When AI changes social service work

Singaporean social workers and social service professionals are over-stretched, often exhausted, and even burnt out. And when AI tools promise increased productivity and reduced administrative workloads – so that workers can focus on the “connective labour” that matters – it is unsurprising that those in the sector are drawn to these tools. Nevertheless, problems arise vis-à-vis the uncritical reliance on bias-susceptible LLMs, shadow AI use, and unintended workload intensification.

When organisations procure AI tools

Very few social service organisations in Singapore will build their own AI tools. Even with current advancements in “vibe coding,” can smaller agencies trust the apps they build or procure? (More on data privacy later.) Among these agencies too, how will they differentiate between tools, and where will they get guidance? Are they able to negotiate with vendors or validate the models?

When practitioners defer to bias-susceptible LLMs

When practitioners are too deferential to AI-generated summaries when assessing service users’ needs and proposing recommendations, they may be tempted to surrender their professional judgement. Such cognitive surrender is even more problematic given the algorithmic bias – including racial bias, stereotyping, and historical correlations – associated with LLMs.

Social workers work with communities who are disadvantaged and marginalised, and these communities are likely to be under-represented in the research and data on which many LLMs are trained. Over time, social workers are not immune to the cognitive offloading disproportionately affecting children and youth, and they will need to navigate LLM inputs in relation to their own training and expertise as well as that of their colleagues and senior social workers.

Relatedly, what happens when service users disagree with AI-assisted summaries endorsed by their social workers? How might they contest those interpretations and prevent them from informing organisational practice?

When sensitive and confidential data enter LLMs

Social service data are highly sensitive as they often relate to people seeking help in moments of vulnerability. Anecdotally, social workers – in addition to Singapore-specific tools like Scribe and Care360 – are using popular LLMs such as ChatGPT and Claude, which have varied privacy policies over whether user data are used for model training. Unmanaged shadow AI use involving identifiable information is a security risk, and the chances of errors and hallucinations notwithstanding, I have been increasingly perturbed by professionals uploading administrative and evaluation spreadsheets for quick data analyses.

Beyond privacy, the sector and its organisations must establish clearer decision rights, in terms of who checks an AI-informed assessment, how overrides are performed, and what accountability looks like if the AI is wrong.

Now, what happens when something goes wrong? We have yet to coalesce around a sector-wide AI policy to answer key questions:

1. What remains a human decision, and what is merely AI-assisted? Answering these questions requires a deeper examination of the value of connective labour and the additive contribution of AI, if any.

2. Within an organisation, which LLMs and AI tools are sanctioned, and for what purposes? An evaluation of the tools in the context of social service roles is needed.

3. How do social workers, alongside their service users, feel about AI use? Unlike corporate entities, where profit and productivity gains take precedence, human-centred interactions and relationships are more ubiquitous within the social service sector. Which leads me to the next threat…

When improved productivity yields only more work

Even if the expected productivity gains from AI tools and LLMs materialise, those gains could in turn intensify workloads and worsen social workers’ already poor work-life balance. Such adverse impact on their well-being could worsen attrition in a sector where burnout has been persistent.

AI may help improve speed and productivity, but it does not fundamentally address the underlying issues of high caseloads – also likely exacerbated by AI complicating helping relationships, as aforementioned – alongside administrative expectations. Pugh noted the following in her book:

“Proponents of AI and apps often seem to focus very narrowly on the interaction, instead of the larger backdrop shaping it, such as why human practitioners have such an impossibly long lists of tasks to start out with.”

In other words, the costs of AI adoption must be weighed. Often underpaid and overworked, social service practitioners may be asked to incorporate LLMs and other tools into their workloads without fully understanding the organisation’s strategy. Pugh elaborated on practitioners’ considerations: “My boss is telling me that I need to start using AI every day and that there’s a high risk that I’m about to lose my job and that maybe there’s going to be permanent unemployment in some job categories. What is your plan for that?”

From an evaluation perspective in the sector, moreover, I am interested in how organisations measure whether AI use has improved productivity outcomes. In fact, how these outcomes are defined reflects organisational priorities. Among workers, do they experience fewer and less intense caseloads? Improved supervision and training? More time away from the workplace? And among service users, how would they benefit from greater AI productivity?

If AI truly saves time, the sector should ask, who exactly gets the time back?

When AI changes the wider sector

As the sector considers the value of human labour in social work as well as the human connection that is unique to the social service sector, it must treat the AI conversation as an extension of the digitisation discourse. In particular, who are the ones who might be left behind (e.g., older adults, those with fewer educational qualifications, lower-income families with lower digital literacy)? Who might need guidance and support (e.g., parents rearing their tech- and AI-exposed children)? And how should practitioners advocate for these groups?

Pugh writes about how AI might worsen existing socio-economic inequalities, with more advantaged groups still benefiting from forms of connective labour, while those less advantaged make do with inferior AI substitutes. She wrote: “Moreover, the overlap of these trends with social inequality means that we are approaching a world in which the affluent pay for personal connective labour from less-advantaged others, who might themselves increasingly have to get it delivered by app or AI. We need to take these trends seriously as a wholesale transformation of connective labour, one that has been taking place without a reckoning that allows us to see the risks involved.”

In Singapore, I expect the following list of threats to grow in the very near future.

When LLMs write the proposal, and LLMs read it too

With funders and grant-making, LLMs have significantly reduced the value and utility of the traditional grant application and proposal process. In summary, we have a deluge of AI-written applications, AI-assisted evaluations, and even AI-generated funding reports. Put otherwise: LLMs write proposals, and LLMs read them too. Like the take-home examination or essay-based assignment in universities, such assessments may no longer test what they once tested.

That means more in-person site visits, clarification interviews and calls, and physical observations of programmes and services.

Share

When fabricated needs undermine genuine ones

Fabricated causes and AI-generated donation appeals could adversely affect trust in a giving ecosystem already facing pressure. While a major Singapore case involving AI-generated beneficiary stories has not emerged, the growing sophistication of deepfakes, impersonation scams, and synthetic media renders this a near-term risk. Singaporeans may also become more sceptical. Conversely, digital platforms with physical verifications or other forms of in-person assurance may gain more traction.

When frontline harms ought to reach policy

Finally, the sector ought to engage with the government and policymaking in relation to the imminent AI divide. Social workers and practitioners, especially based on what they observe first-hand on the ground, are obliged to communicate the harms and threats of widespread AI adoption. They are most likely to encounter these problems and inequalities early and are also best positioned to offer remedies. In this vein, social workers are not just implementers of AI policy.

Some friends and colleagues resent my concession that social workers have no practical choice but to use AI tools, and that some tools are improving documentation and accessibility. I understand and grapple with those concerns. While not completely satisfactory, this list is a starting point – for me, and I hope the sector – to ensure service users’ helping relationships are protected, practitioners are respected and adequately supported, and that connective labour, with all its heart and soul, continues to be available to the most vulnerable in Singapore.

On the face of NVPC’s historical data, the volunteerism and donation figures are weak by recent standards. Donation rates are clearly on a long-run downward trend. Volunteerism is more volatile, but the latest figure remains strikingly low. In this vein, the 2025 NGS is both reporting lower volunteerism and donation rates and reframing what counts as “giving.” In principle, the reframing is defensible. Yet the public report does not provide enough methodological detail or data transparency to adequately support the reframing.

Singapore’s headline volunteerism and donation figures are weak

In 2025, just over one in five people in Singapore volunteered (the 21 per cent figure is a 10-year low). Fewer than half donated money or items in kind (the 45 per cent figure is the lowest since donation data were first tracked in 2004). These figures come from NVPC’s 2025 NGS, which has tracked giving patterns as well as giving facilitators and barriers in the country since 2000, every two to three years.

However, the observations that Singapore’s volunteerism and donation rates are on a downward trend are not immediately clear in the 59-page report. Before these rates are presented on page 17, NVPC first claims that “generosity is widespread.” This claim rests on a broader definition of giving: 68 per cent of survey respondents “engaged in acts of giving” in 2025, which – in addition to volunteering or donating to causes (captured by the aforementioned figures) – includes “everyday acts of mutual aid” or “spontaneous helping” beyond one’s personal networks (e.g., from the report: lifting a stroller, picking up litter, caring for community cats).

My recurring critique of the 2025 NGS is the imprecise presentation and explanation of important statistics. For instance, 68 per cent of people in Singapore reportedly engaged in those acts of mutual aid or spontaneous helping in 2025. “About three in four people living in Singapore have engaged in acts of giving at some point” (emphases mine), and it is this 76 per cent figure NVPC presented in two pie charts, such as the one above, which came without corresponding titles or descriptions. This interpretation is supported by reporting in “The Straits Times,” which states that 68 per cent did such acts in 2025, while 76 per cent recalled doing so at some point in their lives.

It matters that, on the headline indicators, structured volunteerism and donation appear weaker than in recent waves. Consider two sets of figures.

First, as aforementioned, donation rates have been on a long-run downward trend. Volunteerism is more volatile, but the latest figure is weak by recent historical standards. NVPC’s own 2021 report presented the following chart illustrating these patterns. In 2023, the country’s volunteerism and donation rates were 30 and 62 per cent respectively. Such data can motivate important questions:

• Volunteerism-wise, why has it declined since 2016? What explains the ceiling of one-third of Singaporeans who volunteer at least once throughout any given year?

• Donation-wise, how have rates fallen so steadily from the 90s and high-80s? Somewhat counterintuitively, shouldn’t the Internet and smartphones have increased the ease of digital giving?

Second, in addition to volunteerism and donation rates, the reports have also historically reported the number of volunteer hours and individual donation amounts. I created the following table to synthesise the data publicly provided by NVPC. Overall, the median volunteer hours has decreased from 24 hours in 2018 to seven in 2025, whereas the individual donation medians have ranged consistently from S$100 to S$200.

However, consistent with my critique about the imprecise presentation and explanation of important statistics: (a) earlier reports often used averages (or means), while later reports shifted to medians (medians are better, because they are not skewed by heavy volunteers and large donors, yet reporting both means and medians is the best; moreover, when both mean and median are available, such as in 2018, the gap is large, suggesting a highly skewed distribution); (b) the individual donation amounts are not always consistent, depending on whether all donors are included, or if the report only counts those who gave to organisations or through formal means.

The NGS’s revised methodology: Broader definitions, thinner explanations

What also warrants further scrutiny is the methodological approach alongside the level of detail provided of the 2025 NGS. NVPC will likely argue that my fixation on structured volunteerism and donation, thus far, is misguided and overlooks how Singaporeans may “give” generally and informally. In this year’s report (emphases mine):

“Participation in structured forms of giving therefore provides only one part of a larger story. While just over one in five had volunteered in the past year, and nearly half made monetary or in-kind donations, more telling is the pattern within Singapore itself – that much of Singapore’s generosity may flow through informal, relational channels rather than organised structures.”

An NVPC representative added in “The Straits Times”: “If you were to look at it from a layman’s perspective, you may think it’s a drop (in volunteering hours), but in fact, it’s actually a different expression in giving.”

Herein lies my first methodological critique. Focusing on “giving” so generally – including both formal and informal forms – buries the lede about falling formal volunteerism and donation, and the 2025 NGS does not demonstrate much analytical value in comparing “givers” and “non-givers.” In other words, “giving” is such a broad construct that it appears to possess minimal explanatory power.

Of the three substantive chapters, the broadly defined “giving” construct only appears meaningfully in the first. In the second chapter, when describing Singapore’s “participation landscape,” it is formal volunteerism and donation – not the informal acts or behaviours – that remain the units of analysis. We read about, among others, volunteerism and donation rates by age and gender, patterns of volunteerism and donation over a one-year period, as well as how Singaporeans volunteered (e.g., in-person or online, location, proximity) and donated (e.g., money or in-kind, intensity). Relatedly, the formal outcomes are not “giver status” or “forms of giving,” but volunteerism and donation status, patterns, and intensity.

The third chapter details findings with practical implications for understanding what may motivate or deter Singaporeans from societal participation. Here, again, participation largely excludes those widely touted “everyday acts of mutual aid” and “spontaneous helping” in the beginning of the report to focus exclusively on formal volunteerism and donation. More precisely, the 2025 NGS establishes that current volunteers and donors are very useful benchmarks and reference groups.

My second methodological critique is that NVPC has consistently considered informal forms of giving in addition to volunteerism and donation but chose to centre it far more prominently this year. In 2018, it measured “micro-giving,” or “voluntary, spontaneous and everyday acts of giving” (e.g., giving way to others, giving up one’s seat on public transportation, returning one’s tray after a meal). Additionally, “micro-giving” was also treated as an analytic variable vis-à-vis other giving behaviours. In 2021, in addition to “micro acts of giving,” NVPC documented informal individual and community-led giving initiatives, especially in the context of the COVID-19 pandemic. And in 2023, “micro-volunteering” and “micro-donations,” and civic participation were included in the report as “alternative ways of giving.”

I understand the importance of informal giving, particularly as a precursor to formal volunteerism and donation. Having multiple forms of entry – through smaller acts of giving – may make it more likely for Singaporeans to volunteer and donate. Moreover, formal giving can be wrapped up with inequality, wherein socio-economically disadvantaged groups are less likely, on average, to have the time and bandwidth to volunteer and donate.

Still, and this leads to my third methodological critique, if NVPC wishes to broaden the Singaporean conceptualisation and understanding of “giving,” its ambition in the 2025 NGS is not matched, in the public report, by methodological disclosure and data detail. Or rather, it has not provided enough information and wants readers to accept that “the 2025 numbers cannot be compared with those in previous years as researchers used different methods,” even though the report has not made a convincing case.

The request is not unreasonable, and it is not resolved by NVPC saying that their Knowledge & Insights team can be reached for more information. Consider, for example, the International Labour Organisation’s 2011 manual, which the 2025 NGS itself references and cites. The manual published the actual question wording, prompts about unpaid work, the number of hours, and whether the activity was done through an organisation.

The move from a definition-led approach to a criteria-led one is a significant one. As such, NVPC should provide more information, including examples from the survey or even key questionnaire excerpts. Notably, alongside definitions of volunteerism and donation, past versions of the NGS provided important survey questions in their entirety. As recently as 2023, volunteerism was defined as such:

“Volunteering is defined as an activity one does out of one’s own free will, without expecting financial payment, to help others outside of one’s household, family, relative or friends. Volunteering excludes compulsory community work such as [VIA], [CIP] in schools and [CWO], unless compulsory volunteer hours were exceeded. ‘Formal volunteering’ refers to volunteering with registered organisations, whereas ‘informal volunteering’ refers to helping others directly, without going through any registered organisations.”

Based on such a definition, in previous NGS waves, respondents would first indicate whether they have done so, before subsequent questions about frequency and the like. In 2025, such a definition-led approach was replaced by a criteria-led one:

“Rather than presenting respondents with a definition to interpret and apply to their own behaviours, the redesigned questionnaire follows an activity-focused sequence. Respondents are first prompted to recall whether they engaged in any acts of helping others or contributing to society – without labels or definitions being introduced at this stage. For each reported activity, structured follow-up questions capture what was done, who was helped, and the conditions surrounding participation. Activities are subsequently classified as volunteering, donation, or other forms of helping during analysis, based on pre-specified operational criteria – shifting the cognitive burden of classification from the respondent to the research instrument.”

NVPC’s justification to change its approach makes sense on the surface. Notwithstanding my two preceding critiques that the “giving” construct has minimal analytical value and that both formal volunteerism and donation seem to be comparatively more useful, the three-page technical notes section at the end of the 2025 report leaves multiple methodological questions unanswered:

1. Data collection: How were respondents prompted to recall acts of helping or contribution? What were the main and structured follow-up questions? What were the examples of the reported activities?

2. Data analysis: As part of the “pre-specified operational criteria,” the researchers considered activities (a) beyond one’s personal network; (b) that were voluntary and free from compulsion; and (c) substantively uncompensated. Who did the classification, and – if multiple researchers were involved – how were differences resolved? Was inter-rater reliability captured to assess how consistently researchers distinguished between formal and informal giving and whether an activity constituted volunteerism?

3. Constructs and measures: Given the move from a definition-led approach to a criteria-led one, did NVPC consider including both the old and new measures in the same wave, so as to test if the different approaches resulted in different giving or volunteerism rates being derived? Does NVPC have a bridge estimate showing what the 2025 volunteerism and donation rates would have been under the previous definition-led approach?

To NVPC’s credit, the 2025 NGS included a larger sample (3,699 respondents) and reported regression and sequence analyses. My underlying concern is that the public report is not sufficiently auditable.

Subscribe now

Volunteerism and donation rates: Show the data, not just the narrative

My final methodological critique is also a practical suggestion for NVPC to publish more data. Specifically, the 2025 NGS looks aesthetically great, yet the amount of data and information is comparatively scant.

The suggestion to go back to basics can sound trite, yet NVPC’s past reports – even with similar critiques – produced useful insights. Since its inaugural benchmarking survey in 2000, NVPC has collected important demographic data: race/ethnicity, age group, marital status, monthly household income, housing type, employment status, occupation, education level. Volunteerism and donation rates may vary by socio-economic status, and organisations offering volunteer opportunities could better tailor how they recruit and engage potential volunteers.

Against this context, I preferred how NVPC previously presented its NGS findings. Refer above to the collation of screenshots from past NGS reports. The two diagrams on the left are from the 2000 report, presenting that year’s volunteerism rate by age group and monthly household income. The two diagrams on the right are from 2012 and 2016 respectively. They show the overall volunteerism rate in the first row of the table, before presenting the breakdowns across gender, race/ethnicity, age group, and housing type. Additionally, year-on-year statistically significant comparisons are represented by the arrows.

My perspectives are inevitably coloured by my work as a social service researcher. This first means that I’m always in favour of more data and more methodological information. Furthermore, the NGS’s volunteer and donation time-series data – that is, a sequence of data collected about the Singapore population over a period of time – can inform knowledge generation.

For example, I’ve been interested in youth volunteerism for the past decade. Think school-based frameworks such as the Community Involvement Programme and Values in Action, service-learning in the community, and overseas voluntourism. Most recently, I’ve studied and published about the “bathtub volunteerism” phenomenon, wherein young Singaporeans disengage from volunteerism to prioritise their careers and families, only to volunteer again after their mid-40s.

And that journal article referenced NGS data. Specifically, below, I show a chart and table comparing volunteerism rates across six age groups. Across the years, Singaporeans aged 25-34 are among the least likely to volunteer, especially compared with youth and those in mid-life. In 2023, only 23 per cent of them did so, which is below the national average of 30 per cent, and only ahead of those aged 65 and above.

In other words, share more data! To produce a report that is equal parts aesthetically pleasing, accessible to the public, and comprehensive enough for research analysis may be too much to ask of a single publication. If so, given the NGS’s importance and usefulness, perhaps a supplementary publication with all the data, methodological, and statistical trappings is needed. As it stands, the technical notes in the 2025 report are unfortunately inadequate.

Informal giving is important. NVPC is also right that not only structured volunteerism and donation matter, because everyday acts of “giving” and care could form the foundation for more sustained civic participation. However, I cannot accept the narrative from NVPC and the 2025 about a broader story of “giving” at the same time that headline volunteerism and donation rates appear weak. I’d like to see the questionnaire and the fuller sub-group tables. Readers will benefit from information about the coding process and bridge estimates. Otherwise, the language of “giving” risks doing more rhetorical than analytical work, especially when we, as readers, do not have sufficient data to test the narrative for ourselves.

Because Singapore will use AI, the critical questions are: Who benefits? Who bears the risks? And do Singaporeans trust the bargain being offered?

The government is not blind to Singaporeans’ AI anxieties. During his rally speech last week, PM Wong said: “We want to support more of such job transformations. But I know not everyone feels ready. Many Singaporeans are anxious about AI. They ask: Will it replace jobs? Will it be harder to keep up? Will the next generation still have good opportunities? These concerns are real.” Critically, the speech made clear that Singapore’s AI strategy encompasses both technological and social considerations.

So, how worried is the average Singaporean? What do they think about AI? And how effectively do they use AI tools like LLMs (large language models)? Fundamentally, understanding the level of public support will determine whether and how quickly Singaporeans buy into the government’s vision.

For instance, in the US – which houses most frontier models and where the AI technological frontier is relentlessly pushed – Americans are increasingly souring on the technology. This has manifested in protests against AI data centres, and, worryingly, threats and acts of violence, including alleged attacks on property associated with Sam Altman, the CEO of OpenAI (the company behind ChatGPT). Growing American cynicism of AI coalesces around three related explanations.

First, they think the risks outweigh the benefits. AI itself is not an unalloyed good. From the unauthorised use of copyrighted content to environmental and water concerns, Americans are wary: “Half of US adults say they are more concerned than excited about AI, compared with only one in 10 who are more excited than concerned.”

Second, they think the benefits do not flow fairly to them. They worry about being left behind and fear that AI will be disruptive at their expense. People are told by industry leaders that AI will take away their jobs, and against a background of widening income and wealth inequality, those who are socio-economically disadvantaged risk being left further behind.

In 2025, for example, Anthropic CEO Dario Amodei (the company behind Claude) warned that 50 per cent of all entry-level white-collar jobs could be eliminated by AI. Such statements have sharpened anxieties about entry-level positions and labour’s bargaining power. Relatedly, a company (or those who own capital) may benefit from their employees using AI productively, but the workers themselves (or those who produce labour) may not enjoy proportionate benefits.

Third, they think the government is acting too slowly. In general, the regulatory response has been tepid, especially when juxtaposed with the increasing ubiquity of AI tools such as LLMs. A cynical and vociferous public conversant in AI critiques and wary of the unequal accrual of benefits will start to wonder about a government that is missing in action.

What Singaporeans think of AI

Singapore is not the US. Singapore is also not shrouded in the same US political malaise. However, there is little reason to believe that such scepticism or even cynicism will not manifest in Singapore if AI progress and adoption – alongside lagging regulation and inequality threats – are not adequately interrogated. More fundamentally, we know less than we should about how Singaporeans understand, use, trust, and worry about AI.

It is no coincidence that labour chief Ng Chee Meng, during his most recent parliamentary motion on AI “with no jobless growth,” advocated for “a market intelligence and foresight system.” He stressed: “Good research can also help us avoid reactive policymaking.”

To aggregate what we currently know about public perceptions in Singapore, I drew primarily from three large-scale surveys: an Institute of Policy Studies (IPS) report with 3,713 Singapore citizens and PRs aged 21 and above (with a section on AI chatbots), a S. Rajaratnam School of International Studies (RSIS) paper involving 1,014 valid responses from adult Singapore residents (explicitly focused on AI public opinion), and a global KPMG study (which includes a Singapore section).

Importantly, these surveys were completed before PM Wong’s 2026 Budget speech. RSIS’s fieldwork was earlier, in May 2024, while IPS’s was conducted in October to November 2025. They therefore capture public attitudes before the government’s most recent AI push and before this year’s wider discourse about AI agents and accelerating disruption. For example, the viral online essay, “Something big is happening,” captured this wider discourse about accelerating AI disruption in the US.

In summary, what do we know?

Individuals are using AI tools. Companies and organisations, less so: Among individual users, there is some evidence of widespread AI adoption in Singapore, at least relative to the rest of the world, through LLM subscription and usage numbers. In the IPS report, 64 per cent of Singaporeans use AI chatbots. Additionally, there is broad familiarity and engagement with AI tools in life and at work. Conversely, among companies, the Ministry of Manpower (MOM) just reported that 71.5 per cent of firms have yet to adopt AI, 28.5 per cent have started, and only 3.8 per cent are integrating AI into core processes.

Individual AI usage, while widespread, frequent, and often perceived positively, appears limited to practical, informational, and assistive tasks: In the IPS report, the 64 per cent of Singaporeans who use AI chatbots are mostly searching for information, reviews, or recommendations, getting assistance with school or work tasks, or planning and organising events or trips.

“School or work tasks” is an admittedly broad category. As such, the KPMG study offers some workplace-related indicators. A necessary caveat here is that KPMG’s study respondents appear far more AI-exposed than the average firm in MOM’s establishment survey: 78 per cent said their organisation uses AI (compared to the earlier MOM finding that only 30 per cent of Singaporean firms have adopted AI). Overall, they have grown more reliant on AI. Positively, the usage and reliance have resulted in increased efficiency, quality of work, innovation, and revenue-generating activity. Negatively, a smaller proportion reported increased workload, stress, and pressure.

Singaporeans are cautiously optimistic and somewhat clear-eyed about AI’s limitations: A topline finding from the IPS report is that 92.8 per cent of respondents agreed that “People need to exercise caution when using AI chatbots.” In fact, they believe that these chatbots can share misleading information (87.3 per cent agreement), give harmful advice (75.3 per cent), create unrealistic expectations about human relationships (74.8 per cent), make it harder for people to form social connections with others (73.1 per cent), and make individuals less likely to seek help from real people (72.4 per cent).

This is echoed in the KPMG study, in which 65 per cent of respondents reported experiencing or observing loss of human interaction and connection associated with AI. While only 25 per cent believe that AI risks outweigh the benefits, 81 per cent are concerned about AI’s negative outcomes, and 49 per cent have personally experienced or observed negative outcomes from AI.

The RSIS paper asked its respondents the different ways AI could be used for good and harm. The most common benefits included increased efficiency and productivity, enhanced quality of life, language and learning, and assistance in planning and decision-making. On the other hand, the most common challenges included cybersecurity threats and negative societal impacts. Although these categories are very broadly defined, Singaporean perceptions of good and harm could be characterised as nuanced.

The gap between AI harms and AI regulation

Cumulatively, given their awareness of AI’s potential harms and desire for a fair share of AI’s benefits, Singaporeans want some AI regulation.

Globally, consider the common critiques of LLMs and AI image generators. That LLMs are built upon and use datasets scraped from the web, often bypassing creator consent, proper attribution, and financial compensation. That growing electricity demands and cooling water required to operate global data centres can be harmful to the environment. That deepfakes and AI misinformation disproportionately target and harm marginalised groups, especially women and minority communities. That online child safety is paramount, with growing knowledge that smartphones and social media have been deleterious for positive youth development.

The Singapore government has not been absent from AI governance, but much of the response feels uneven. In my opinion, regulation around labour displacement, model transparency, copyright, and environmental externalities remains disproportionately less developed than the accelerated pace of AI promotion.

What might regulation look like, particularly of immediate AI harms? Expect discourse about data centres in Singapore and South East Asia (especially Johor, Malaysia) and the expected environmental costs. In parliament, MPs are raising questions about how perpetrators of technology-facilitated sexual violence (e.g., AI deepfakes and image-based sexual exploitation) may be punished and how victims may seek remedies. In schools, there are concerns over AI-based cyberbullying as well as the learning and developmental impact of AI usage, and the Ministry of Education is similarly trying to balance AI benefits and harms.

If we accept that AI progress is a fait accompli, and the Singapore government believes that Singaporeans have to adapt or lose competitiveness, AI promotion and regulation are not necessarily antithetical to each other. In fact, regulation may be what sustains the public’s continued consent for AI promotion. In the KPMG study, 67 per cent of respondents believe that AI regulation is required, and 52 per cent believe current safeguards are sufficient.

Among the multiple problems and threats that AI might bring about, the government and Parliament have paid the most attention to economic and employment disruption. To its credit, the government says Singaporean workers will not be left behind. PM Wong has said: “We will exploit AI to grow the economy, and we will ensure that growth translates into good jobs and better wages.”

If the long-term aim of AI advancement and adoption is increased productivity alongside higher-paying and better-quality jobs, the government’s overall approach – such as job matching and institutional restructuring – seems to be that the workers of today will be protected through skills training as well as career conversion and job reskilling programmes.

Subscribe now

I see two issues with that approach. First, assuming there are indeed productivity gains in some sectors, at what cost to the worker? Rather than reducing workers’ work hours or improving their work-life balance, LLMs and AI tools may instead increase productivity expectations and stressors. One suggestive study of a US-based tech company with 200 employees has provided some evidence of labour intensification. In the beginning, employees worked faster, took on more work, and worked more hours, yet it was unsustainable, because the initial productivity could give way to “cognitive fatigue, burnout, and weakened decision-making.”

Second, if we take seriously the prognostications of AI industry leaders about widespread job loss, then what careers are Singaporeans supposed to be converting to, or what jobs should they be reskilling for? This uncertainty matters most for fresh graduates and younger workers, whose first jobs include entry-level responsibilities that may be displaced by AI tools and LLMs.

In other words, what comes after job displacement is not clear. Moreover, the track record of training and reskilling programmes is neither clear nor uniform. Previously, software engineering and coding-related industries were the purported solutions to keep up with technological disruption. Now, the tech sector is threatened.

An emerging AI divide

If these disruptions materialise and persist, the threat of socio-economic inequality is likely to be most pernicious. Many are worried that the economic benefits from AI advancements will primarily accrue to a small group of elite corporations and wealthy individuals. Past evidence has shown that technology-facilitated economic disruption disadvantages those who are already disadvantaged.

Take the aforementioned IPS report as an example. Degree-holders are using AI chatbots much more (at 81.4 per cent) than those with post-secondary education (at 65.6 per cent), and only 30.9 per cent of those with a secondary education and below use the chatbots. Those who are younger are also more likely to use the chatbots. The RSIS paper also found that one’s education level was associated with whether they were likely to recognise AI’s benefits and harms, with warnings about uneven familiarity and existing inequalities.

Put otherwise: how these different groups use AI and chatbots is likely to vary, and a continued digital divide is expected. In this vein, the AI conversation is an extension of the past digitisation discourse. What happens to those who are left behind?

I fully understand that the uncertainty associated with AI and its rapid development renders policymaking more challenging. At the same time, the government finds itself in somewhat of a pickle because it is going all in on AI against Singapore’s context of income inequality and wealth inequality (in a paper published days before the 2026 Budget). More precisely, it does not bode well for PM Wong’s vision of a “We First” society if some households feel that their prospects for upward social mobility are circumscribed, and they similarly believe that AI will only widen those existing gaps.

Ultimately, the question is not whether Singapore should use AI. It will, and it arguably should. The harder question, I think, is whether the AI transition can be made politically and socially legitimate. A “We First” Singapore cannot be built without serious discussions about fairness and equality. Regulation must catch up with immediate harms, and Singaporeans must believe that AI-related productivity gains are not limited only to higher expectations for workers or higher returns to capital.

If AI is truly to become a national advantage, Singaporeans must be convinced that it will not become another engine of unequal opportunity.

It’s 8pm on May 3 in Singapore, which marks the conclusion of Polling Day (and GE2025). The nine-day campaign has whizzed by, and I think the election will ultimately be decided by the following five numerical figures.

#01: The approximately 30 per cent of undecided voters. Online market research firm YouGov Surveys – using data collected a month ago – found that 19 per cent of voters were undecided, and 11 per cent would decide during the campaign period (with three per cent more deciding at the voting booth). More recently (April 15), and similarly, market research company Blackbox Research said 30 per cent of voters were still undecided. Undecided voters appear to be younger. A June 2024 voter sentiment survey by magazine “Jom” also documented that younger Singaporeans were less confident of their vote.

#02: Nine per cent (GST) versus $5,000 (financial support): Obvious prognostications about the centrality of bread-and-butter issues – specifically, the cost of living, housing affordability, as well as jobs and salaries – have borne out, as expected. For the undecided voter, a vote against the ruling People’s Action Party (PAP) will likely reflect dissatisfaction with their socio-economic conditions and that of Singapore, epitomised by the opposition’s persistent focus on the nine per cent GST hike. Conversely, a vote for the PAP may indicate agreement with the government’s justification for the increase, alongside the perceived adequacy of financial support (e.g., the prime minister repeating how the average Singaporean family with two young children will receive about $5,000 in comprehensive financial support this year).

#03: For the PAP, it helps that it’s 2025, not 2024. Last year was a particularly terrible electoral year for incumbents around the world. This year, Singaporeans appear more satisfied and optimistic, and the experiences of some voters are backed by objective indicators such as rising household incomes, lower income equality, and fewer households requiring financial assistance. Based on the aforementioned survey from Blackbox Research, overall, 88 per cent of the respondents believe that Singapore is headed in the right direction. Year-on-year government satisfaction on economic issues has further improved, which can only bode well for the PAP:

#04: Separately, beyond bread-and-butter issues, many Singaporeans may have another numerical figure in mind: Two-thirds, or the supermajority the PAP possesses in parliament (since pre-independence), giving it the ability to amend most articles in the constitution. For further reference, in GE2020, it won 89 per cent of all parliamentary seats. The closing arguments of most opposition parties have centred on denying the ruling government of that supermajority, by increasing parliamentary diversity across parties. Polymarket – a cryptocurrency-based prediction market deemed illegal in Singapore – still indicates that the PAP will secure at least 78 seats (as of 1pm on May 3 in Singapore, when I finalised this newsletter), translating to at least 80 per cent of all seats. However, even if the loss of a PAP supermajority is very unlikely, the argument will feature in voting considerations.

In a post-GE2020 survey by the Institute of Policy Studies, “pluralist” voters – “who agree that there is a need for change in the electoral system, checks and balances and different voices” – increased from 18 to 22.4 per cent between 2015 and 2020.

#05: Finally, will it be one or many opposition parties? The Workers’ Party (WP) has dominated as Singapore’s main opposition party in the past two decades, and GE2025 will determine if the other parties are resigned to be “coalitions of electoral losers.” WP’s non-contestation of Marine Parade-Braddell Heights GRC quickly became a non-issue, and the overall conduct of the nine-day campaign – such as the parties’ online presence and content, rallies, as well as on-the-ground mobilisation – has set the WP, Progress Singapore Party, and Singapore Democratic Party further apart from the rest of the pack. Finally, the diversity of match-ups will offer clues of the proportion of anti-PAP votes vis-à-vis pro-opposition party ones.

We’re at the halfway mark of the GE2025 campaign, and – unsurprisingly – the discourse remains centred on bread-and-butter issues (especially the cost of living), immigration, and the ruling People’s Action Party (PAP). As such, I’m sharing three socialservice.sg podcast episodes focused on these issues, highlighting key insights and excerpts:

1. “It’s the economy, stupid?”: Inflation, incomes, inequality, and Singapore’s general elections (with Dr. Linda Lim)

2. Singapore, first? Issues of migration, citizenship, multiculturalism, and nationalism in Singapore (with Dr. Rebecca Tan)

3. Looking back and looking ahead: Dr. Kenneth Paul Tan on PAP’s governing principles, PM Lawrence Wong, and the next general election

“It’s the economy, stupid?”

Inflation, incomes, inequality, and Singapore’s general elections (with Dr. Linda Lim)

How do the economy and economic conditions influence electoral outcomes, both broadly and in Singapore? What can we understand from Singapore’s ranking as one of the world’s most expensive cities as well as the country’s Consumer Price Index, or the CPI? And if the fundamental root cause of Singapore’s low wages can be attributed to widespread cheap sourcing and the influx of foreign workers from very low-wage neighbouring countries, how do we address that problem?

To answer these questions and more, we host Dr. Linda Lim, Professor Emerita of Corporate Strategy and International Business at the University of Michigan. We previously hosted Dr. Lim on this podcast in 2020, in an episode focused on the minimum wage, Singapore’s Progressive Wage Model, and low-wage labour in the country.

In GE2025, a contestation between “objective” and “subjective” economic evaluations is also coming into focus. Dr. Lim reflected on this, in the episode:

“In Singapore, the problem is that the inequality people perceive and experience is not captured in statistics. We see increased numbers of rich people, especially foreigners, throwing their money around, buying whole floors of luxury [condominiums], fleets of luxury cars, luxury retail, while housing prices climb beyond my affordable range. My food and utility bills, hawker stall prices, go up, not to mention GST. I’m just not going to be happy …

I just want to be able to continue consuming what I consume, without it being priced out of my range, without having to constantly downsize. I don’t want to drive a Lamborghini. I just want a Grab ride to be affordable. I don’t want to eat at fancy high-end restaurants. I just want my chicken rice to have the same amount of chicken it had last week. I don’t want to live in Ridout Road or any GCB. I just want to be able to afford an HDB flat. I also want to be able to meet my aspirations for higher consumption in the future.”

Listen to the episode with Dr. Lim.

Singapore, first?

Issues of migration, citizenship, multiculturalism, and nationalism in Singapore (with Dr. Rebecca Tan)

Issues of migration, citizenship, multiculturalism, and nationalism have featured prominently in both our politics and elections. With Dr. Rebecca Tan of the National University of Singapore, we take a deep dive into these issues. Fundamentally, she asks, what is fair and what is right? And how do we treat individuals more equitably?

Specifically, we zoomed into five key events in Singapore’s past decade: First, GE2011; second, the 2013 Population White Paper; third, the 2013 Little India Riot; fourth, the focus on the 10-million population figure during and after GE2020; and fifth, COVID-19 and the plight of migrant workers. After analysing these events, Dr. Tan highlights data challenges and how all the issues – cumulatively – could feature in the next general election.

In GE2025, while immigration has not featured as prominently as cost-of-living issues, references to the “foreigner” or “foreign talent” have loomed in the background:

“What I find interesting is the framing of what counts as Singaporean-ness, who is a good Singaporean, who is deserving of citizenship and belonging, and all the benefits that come from being of the Singaporean community … To see, when they talk about foreigners, especially the opposition parties, how do they draw a distinction between natives or locals and the foreigner? How do they construct who’s a good Singaporean and who’s not?

We’re going to see that in terms of public discourse, whether it’s a numerical one, but also kind of a more value-laden one about Singaporean-ness, protecting the Singaporean nation, and so forth.”

Listen to the episode with Dr. Tan.

Looking back and looking ahead

Dr. Kenneth Paul Tan on PAP’s governing principles, PM Lawrence Wong, and the next general election

On May 15, 2024, at the Istana, Singapore’s fourth prime minister Lawrence Wong took his oath of office and was officially sworn in, taking over from prime minister Lee Hsien Loong. This political transition is significant for a host of different reasons, and so we take a deeper dive with Dr. Kenneth Paul Tan of Hong Kong Baptist University. Specifically, we considered three potential electoral scenarios for Mr. Wong and his ruling PAP, and asked Dr. Tan, is it the PAP’s election to lose, or the opposition parties’ election to win?

These were the three scenarios we explored:

1. Most optimistically for them, Mr. Wong and the PAP do much better than they did in GE2020, scoring scores five to 10 percentage points higher than they did in 2020?

2. They do about the same as they did during GE2020, losing a small handful of seats, a GRC here, maybe two or three SMCs there, but still maintaining their supermajority in parliament.

3. Finally they do much worse. The PAP is close to losing or even loses its parliamentary supermajority.

Additionally, in a recent journal article (open access), Dr. Tan further detailed “a series of high-profile scandals in recent years have eroded the PAP’s long-standing reputation for incorruptibility.” The scandals include the Ridout Road saga, corruption case involving former transport minister S. Iswaran, handling of TraceTogether data, SimplyGo initiative, attempted NTUC Income-Allianz deal to privatise the former, ACRA-NRIC lapse, as well as SingPost saga. Dr. Tan added:

“These examples illustrate how the PAP government’s deep entwinement with key institutions and policies means that operational failures, lapses in transparency, and public dissatisfaction often reflect directly on its leadership. Collectively, these incidents set the stage for the upcoming 2025 general elections, which will serve as a critical test of the PAP’s ability to navigate its evolving legitimacy crisis …

At least two narrative strategies are likely to dominate the coming election. First, opposition parties will leverage public discontent over PAP’s fourth generation leadership and recent scandals to position themselves as a credible and necessary check on the government and an alternative to PAP’s technocratic elitism. Second, the PAP is likely to reinforce its traditional narratives of national security, economic stability, meritocracy, multiracialism, and pragmatism, while leveraging its significant influence over mainstream media to vilify critics and political rivals, ensuring its narratives continue to dominate public discourse. It will emphasize the dangers of political experimentation, particularly in an era marked by global uncertainty. However, these PAP narratives may struggle to resonate with a populace that increasingly perceives them as outdated and disconnected from contemporary realities.”

Listen to the episode with Dr. Tan.

Across the first days of GE2025, all political parties unanimously agree on one thing – that a trade- and globalisation-dependent Singapore is vulnerable, as tensions between China and the United States (US) intensify.

Where the parties differ – based on the opening salvos launched by party leaders – are their assessments of how Singaporeans have fared in the past five years, as well as their visions of what the Singapore parliament and government should look like, so that the country is best-positioned to meet those looming international threats.

For the ruling People’s Action Party (PAP), given the wide-ranging and adverse effects of the COVID-19 pandemic, it hasn’t necessarily been “the best of times” since 2020. However, prime minister Lawrence Wong is asking voters to keep their faith, to both look back and take a longer view, respectively, of the internationally recognised success of “the Singapore story” and the PAP’s plans for the future. On the other hand, with inflationary pressures and the rising cost of living, the opposition parties are arguing that it’s truly been “the worst of times” for the average Singaporean.

The Singapore Democratic Party (SDP) is unequivocal that the PAP and its ill-advised policies are to blame. That the ruling party has gone wrong and its ministers are oblivious to Singaporeans’ everyday struggles. The Progress Singapore Party (PSP) and Workers’ Party (WP) don’t disagree with the SDP. Yet they explicate a more nuanced message – perhaps unsurprisingly, given their parliamentary representation and track record in parliament – that they’ll definitely not form the next government. Instead, in a PAP-dominated 15th parliament, diverse opposition MPs, as a “loyal opposition,” are needed to counter “groupthink.”

Frustratingly, because it has been an electoral tale of these two general versions of Singapore, any substantive discourse on policies has mostly stopped at: “Read our ostensibly detailed manifestoes!” Thus far, party leaders seem content to talk over each other, rather than with one another. For me, in this remaining week of campaigning, the heavier responsibility falls on Mr. Wong and his PAP to set the engagement terms over specific socio-economic issues and policies of interest. The party has introduced dozens of new candidates, some of whom could be political office holders. Can they speak more substantively about issues and policies? Or take up challenges to debate policy?

Furthermore, if we accept Mr. Wong’s premise that a sizeable political opposition is determined, here to stay, and desires growth, I’d like to hear him share, from the PAP’s perspective, the role of the opposition in parliament.

“It was the best of times”

To start, Mr. Wong’s party political broadcast on April 25 was largely similar to his April 24 kick-off speech, during the PAP rally for Marsiling-Yew Tee GRC. Delivering a standard stump speech is part and parcel of most political campaigns globally, though his is becoming a tad too templatised and too formulaic. His speeches start with wary proclamations of a dangerous world, before turning to the PAP’s historically strong and stable stewardship. “This election comes at a time of profound global change. A storm is coming upon us,” the prime minister warned, and he emphasised that these were not exaggerations or attempts at fear-mongering.

With tariffs and trade barriers, the world is becoming more turbulent. The risk of a recession cannot be ruled out, Mr. Wong added. It is not the PAP versus the opposition, but the little red dot that is Singapore versus a world that is “dangerous,” “troubled,” “contested,” and “fractured.” Therefore, it is against this tumultuous global background that the prime minister explains his need for a good team to not only continue fighting fires, but also plan ahead by reviewing and updating policies.

That the PAP would seek to capitalise on a rally-round-the-flag effect was predictable for many, including Prof. Cherian George of the Hong Kong Baptist University. He wrote, at the start of GE2025:

“In ordinary times, it is hard for the government to educate Singaporeans about Singapore’s vulnerabilities in a global economy. Donald Trump is doing the PAP a favour. This year, there is no need for the PAP to explain supply chains, interest rates, and the invisible forces of global competition. The havoc being unleashed by the most powerful man in the world is as visible as Godzilla rising out of the Singapore Strait in some monster movie.

In this environment, the PAP may be exceptionally successful in projecting itself as able and trustworthy guardians of a safe haven in a dangerous world.”

To Mr. Wong’s credit, he’s gone beyond mere references to global threats as well as Singapore and PAP exceptionalism. When acknowledging concerns over the cost of living and employment, his go-to anecdote is Budget 2025-related, that the average Singaporean family with two young children will receive about $5,000 in comprehensive financial support this year. Furthermore, he’s alluded to aspirations for children, youth, and working adults, and to policies on affordable public housing and preventive healthcare.

“It was the worst of times”

Conversely, when you listen to the first two rally speeches made by SDP’s Chee Soon Juan (April 24 and 25), he paints a different, more negative picture of the past five years. Dr. Chee laid out examples demonstrating the government’s “profligate spending,” such as on the Founders’ Memorial. He also criticised the PAP’s insistence to raise the GST – alongside other increased utility, transportation, and healthcare expenses – despite the financial hardship experienced by Singaporeans. “The PAP raises the GST and then tells you that it is giving you vouchers to help you cope with the increased GST,” he added.

Overall, the SDP secretary-general is holding the PAP government responsible. “Here in Singapore, we have the potential, we have the talent, we have the energy to do big things.” He concluded: “The only thing that’s stopping us is the PAP squatting on top of all that happens on this island and hoarding front and centre of everything we do.”

The PSP and WP – more specifically, their party leaders – have taken a slightly dissimilar rhetorical approach. During his party political broadcast, PSP founder and chairman Tan Cheng Bock described PSP’s role as a “loyal opposition that proposes constructive alternative policies.” Given that it is impossible for the party to form the government, they will instead push and help the government by offering diverse and constructive views. Even more concretely, the party takes pains to highlight the respectable performance of its two NCMPs in the 14th parliament.

Additionally, at his kick-off rally speech on April 24, Dr. Tan stressed that the PAP of today “has lost its way,” with the party’s candidates and MPs increaingly susceptible to groupthink.

WP secretary-general Pritam Singh has also echoed Dr. Tan’s notion of serving as a “loyal opposition.” During his party political broadcast, Mr. Singh started with an explicit message that his party will not form the next government. Rather, WP candidates are qualified and viable because they can handle municipal and town council affairs, speak freely on topics in parliament, and help build for the future.

At his kick-off rally speech on April 24, Mr. Singh did not shy away from lambasting the PAP government – and its poor timing – for raising the GST and consequently turbocharging inflation. As aforementioned, his SDP and PSP counterparts have delivered the same message. Nevertheless, Mr. Singh took a step further to describe how “a vote for the WP is your instrument of change.”

He detailed five observations. First, his stress-tested WP – loyal to Singapore and Singaporeans, not to the PAP – has done well in parliament, complementing the ruling government. Second, independent ideas by its MPs help counter groupthink, when they offer alternative ideas for structural policy improvements. Third, its manifesto offers not “motherhood statements,” but a concrete and realistic vision. Fourth, the party pressuring the government as the voice of Singaporeans. And fifth, that his party is working towards and building a more balanced political system.

Are you better off than you were… Five years ago?

At the end of the day, with less than week of campaigning left, how many times can the opposition parties – including and beyond the PSP, SDP, and WP – lambast the ruling government for its perceived failures since 2020? To remind the voters that it’s been the worst of times. And for the PAP, for how long can it rely on the “a PAP-governed Singapore is a safe harbour in a turbulent world” shtick, with promises of the best of times within our collective grasp, before its candidates seriously engage on substantive issues and policies and reckon with the permanence of an opposition (loyal or not!) in our politics.

I have no dog in this fight. For Singapore’s opposition parties, their campaigns are straightforward if and when GE2025 is reduced to these questions uttered by then US presidential candidate Ronald Reagan in 1980, against incumbent president Jimmy Carter: “Are you better off now than you were four years ago? Is it easier for you to go and buy things in the stores than it was four years ago? Is there more or less unemployment in the country than there was four years ago?”

If so, if I were in the shoes of Mr. Wong and his PAP, do I continue pointing to Singapore’s objective economic conditions – higher household incomes and lower income inequality, as well as the country’s inflation growth hitting a four-year low – signalling that things will get better? Or do I take a more impassioned approach with the issues and policies, setting the terms of engagement and debate, articulating a more compelling vision for the next five years – and beyond?

Good morning, and Happy Nomination Day! Across the next nine days of campaigning (April 23 to May 1), at the end of each day, I’ll summarise the most important insights and developments. Because you’re likely to be inundated with news content, I’ll keep the newsletters snappy (hopefully 500 words or fewer!), with a heavier focus on social service and social policy issues.

Yesterday, I reflected on (new) candidates vis-à-vis their legislative aspirations, with hopes of candidate introductions being less banal. Today, before election hustings formally begin, I take a deeper dive into the party manifestoes of the ruling People’s Action Party (PAP) as well as the main opposition parties (i.e., the Workers’ Party, WP, Progress Singapore Party, PSP, and Singapore Democratic Party, SDP).

I had three overall thoughts, though the first may be unpopular. In summary, first, the PAP manifesto is not detailed, but it is entirely fit-for-purpose. Second, all four manifestoes centre bread-and-butter considerations, and the opposition parties have mooted meaningful policy proposals. Yet, to me, the main political contestation will pit “objective” macro-economic conditions against “subjective” Singaporean experiences about income inequality and social mobility. Finally, besides highlighting some social service and social policy proposals of interest, I’m excited to see how the manifestoes will translate into campaign messages across the next nine days, particularly for the opposition parties.

(Very, very usefully, independent journalist Kirsten Han has created a database which extracts policy proposals from the manifestoes of all political parties contesting in GE2025. Users can browse the proposals by theme and party. It’s an excellent resource for the election.)

PAP’s fit-for-purpose manifesto

First, the PAP did not need a more detailed manifesto, because the party can draw from the 2022 Forward SG Exercise (this is explicitly acknowledged in the manifesto), and – more crucially – it is campaigning on the government’s track record (more on that later). In other words, its manifesto is entirely fit-for-purpose. Its 12-word slogan, however, is a garbled mouthful: “Changed world, fresh team, new resolve – Securing a brighter future for you.”

The PAP manifesto consists of two parts. In the first, it referenced its GE2020 slogan – “Our Lives, Our Jobs, Our Future” – while making the case for its five-year track record: “Together we overcame a pandemic, battled inflation, revived our economy and refreshed our social compact.” In the second part, in each of the nine chapters, it specifies how policies will continue for respective groups of Singaporeans. For instance, in Chapter 3 (“For Every Singaporean – The Support That You Need”), lower-income families are promised support through ComLink+, an enhanced, post-Forward SG version of the Community Link (ComLink) scheme.

“Objective” and “subjective” economic evaluations

My second reflection about the four party manifestoes I read – including the PAP’s – is the obvious centrality of bread-and-butter considerations, especially since inflation and the cost of living are top of mind for Singaporean voters. The SDP leads with a focus on fair taxes, specifically a GST reduction, and its manifesto also proposes housing and healthcare policies. The first two sections of WP’s manifesto covers affordability and cost of living concerns as well as economic growth and opportunities. Similarly, the first four chapters of PSP’s manifesto is ordered as such: Cost of living, housing, jobs and wages, and social safety net.

Be that as it may, to me, GE2025 will primarily be contested around the following question: How will the parties juxtapose the “objective” macro-economic conditions (i.e., strong economic indicators) and “subjective” experiences about income inequality and social mobility? The PAP will point to strong data demonstrating Singapore’s economic success (i.e., higher household incomes and lower income inequality, as well as the country’s inflation growth hitting a four-year low). Conversely, the opposition parties will assert the following: If the macro-economic conditions are so good, why do some Singaporeans still feel bad about the economy?

Last year, in our podcast episode with Prof. Linda Lim of the University of Michigan, she explained this socio-economic tension, with obvious implications for political contestation:

“In Singapore, the problem is that the inequality people perceive and experience is not captured in statistics. We see increased numbers of rich people, especially foreigners, throwing their money around, buying whole floors of luxury [condominiums], fleets of luxury cars, luxury retail, while housing prices climb beyond my affordable range. My food and utility bills, hawker stall prices, go up, not to mention GST. I’m just not going to be happy …

I just want to be able to continue consuming what I consume, without it being priced out of my range, without having to constantly downsize. I don’t want to drive a Lamborghini. I just want a Grab ride to be affordable. I don’t want to eat at fancy high-end restaurants. I just want my chicken rice to have the same amount of chicken it had last week. I don’t want to live in Ridout Road or any GCB. I just want to be able to afford an HDB flat. I also want to be able to meet my aspirations for higher consumption in the future.”

How will the opposition parties make their case? You see it peppered throughout their manifestoes. They include arguments about how government taxes and transfers may have improved income inequality over time, but Singapore still compares poorly to other wealthy countries. There are references to wealth distribution and inequality. The SDP and PSP, in particular, emphasise fairness and dignity throughout their manifestoes.

How will party manifestoes shape the nine-day campaign?

Third, and finally, the burden was never on the opposition parties to publish extremely detailed policy proposals. In fact, generally and substantively, the manifestoes are very thoughtful. The parties and their candidates do not benefit from incumbency, and not being in power also means they do not have the resources of the civil service at their disposal. Even so, beyond the GST-related proposals, I spy a few recurring and interesting themes in relation to social service and social policy issues:

1. Strengthening existing social safety nets, including and beyond ComCare (for low-income Singaporeans) and Silver Support (for seniors). Interestingly, the PSP also calls on the government to “simplify the existing multiple overlapping support scheme system to a few national schemes.”

2. Calls for a national minimum or living wage, a policy which gained traction during GE2020 (and also, in my opinion, following the Minimum Income Standard study). The WP, PSP, and SDP are advocating for it, albeit at different wage levels. Relatedly, the WP reiterates its call for a national poverty line.

3. There are proposed policy intersections between poverty or inequality and education, with recommendations in the manifestoes to better support children from low-income families through tuition programmes (WP) as well as pre-schools and universities (SDP).

4. Finally, the WP advocates for a wealth tax: “Impose a net wealth tax of between 0.5 and 2 per cent on the assets of the top 1 per cent of wealth holders in the country.”

Evaluations of how these opposition party manifestoes will translate into campaign messages – as we get further away from the tangible policy proposals – are probably the most subjective (which is also why I’ve left this to the end). The most systematic way I could organise my reading notes was based on the party’s proximity to parliamentary representation.

• The SDP doesn’t have an MP or NCMP, and thus its slogan, “Thrive, not just survive” reads like an aspirational rallying cry against the PAP, which has prioritised the country’s economic growth.

• With its two NCMPs, the PSP’s “Progress for All” slogan is directly supported by its members’ active contributions in parliament, so as to dispense with the lazy narrative that opposition parties do not suggest alternatives.

• The WP, like the PAP, includes in its manifesto a chapter of its ostensible impact on public policy. As the largest opposition party, its approach feels like a combination of the SDP’s and PSP’s, to marshal more voters to its cause, demonstrate its legislative contributions, and make the case that additional political diversity is much needed.

What’s more likely, I think, is that the opposition parties will wait for the PAP to stake its electoral campaign position more clearly, before framing and fine-tuning their messages in response.

“It’s up to us MPs to make sure that peoples’ voices are heard in parliament”

On a final, sad note, former MP Louis Ng – the most active parliamentarian in the 14th term of parliament – announced that he will not be contesting in GE2025. The two-term MP is stepping away for family reasons, and despite interventions by Law and Home Affairs minister K Shanmugam and prime minister Lawrence Wong.

By all accounts, Mr. Ng’s presence will be missed by his residents. His voice will also be missed in parliament, including by social workers. We’ve hosted him twice on the socialservice.sg podcast, and both episodes highlighted just how seriously he took his responsibilities. In January 2022, as a leading legislative voice on the work and welfare of the Singaporean social worker, he shared about his public consultation initiative, which covered emerging issues around pay and compensation, burnout and retention, case management and caseload ratio, as well as community work.

And in July 2023, as his active parliamentary engagement gained public attention and traction, we quizzed Mr. Ng about how he engaged legislative assistants – hired using an MP allowance – to help him with parliamentary questions and quizzes. He gamely shared about his motivations, the process behind drafts and re-drafts, and how he and his team integrated feedback from walkabouts. In that spirit, I leave you with the following quote from the episode:

Much has been made about the banality or unoriginality of GE2025 candidate introductions – especially those from the ruling People’s Action Party (PAP) – when candidates turn to sob stories about their humble beginnings, difficult childhoods, and challenging backgrounds. Predictably, their stories conclude with an inevitable triumph against the odds, thereby demonstrating Singapore’s meritocratic promise, as if to say: “If I can do it, so can you!”

Strangely too, and perhaps somewhat paradoxically for the ruling party, it has also placed less emphasis on its new candidates’ credentials. It feels paradoxical because, historically, a key PAP selling point has been its ability to attract the crème de la crème across the public, private, and non-profit sectors, to serve at a higher level for the country. To form the government! Instead, on the eve of Nomination Day, the PAP candidates may appear ready to handle municipal issues and relate to the average Singaporean – with a listening ear and a willing heart – but we know much less about their legislative aspirations.

(There is also ongoing discourse about the comparative “calibre” of candidates across political parties.)

Ultimately, what frustrates me most is how little I’ve learnt about these candidates, in their potential roles as parliamentarians. As someone in the social service sector, besides bread-and-butter issues, as I’ve written, I want to know about issues revolving around cohesion, cultivation, caregiving, and compassion. In a recent video explainer, Nanyang Technological University’s Prof. Teo You Yenn argued that factors such as relatability, likeability, and even candidates’ hobbies or pets are potential distractions. Instead, with limited information about candidates, voters should pose these questions:

• What are the challenges you see in Singapore society?

• What issues do you care about?

• How do you intend to prepare for your role as an MP?

Has the PAP made a strategic choice to keep its candidate introductions as safe as possible? After all, senior minister Lee Hsien Loong argued last week that candidates should not be judged just as individuals, but also the party for which they stand. He elaborated: “And also consider carefully that when you vote for an MP, you are not just voting for the person. You are voting for the party which stands behind her or him, and what the party needs by way of support and mandate in order to be able to serve Singapore.”

I’m taking up Mr. Lee’s suggestion to consider both the candidates and their parties. Today, specifically for my constituency first, I briefly study the candidates and their legislative aspirations: How do I learn more about the candidates? Then, more generally, after a broader scan of new candidate introductions, I evaluate: What works with new candidate introductions?

Tomorrow, I take a deeper dive into the party manifestoes of the PAP as well as the main opposition parties (i.e., the Workers’ Party, WP, Progress Singapore Party, PSP, and Singapore Democratic Party, SDP).

How do I learn more about the candidates?

The obvious starting point for any Singaporean voter is to identify the (potential) candidates in your SMC or GRC. For me, I’ll vote – by post! – in Yio Chu Kang SMC. Right now (as of April 22), it’s likely a contest between the incumbent PAP MP Yip Hon Weng and Dr. Michael Fang of the People’s Alliance for Reform (PAR).

In many ways, this match-up is probably emblematic of the frustration felt by many voters, whose constituencies are not contested by one of the main opposition parties. In a healthy democracy, I value political contestation. I want diverse views to be represented robustly. Furthermore, I would never belittle the efforts of anyone willing to put their name forward, because it is a significant commitment, and because no walkovers are good.

But, my goodness, do Dr. Fang and the PAR make it very difficult.

Dr. Fang doesn’t have an official online website. The PAR website is an absolute mess. I couldn’t find any of his social media profiles or online footprint. Hence, all I know about him vis-à-vis GE2025 – that he is a 48-year-old health sciences lecturer who contested in GE2020 under the Peoples Voice banner – is from a brief report in “The Straits Times.” Besides a handful of pre-2025 content, there is only a 44-second “CNA” interview clip of Dr. Fang describing himself as “a barometer for change” (What does that even mean?)

Notwithstanding the benefits of incumbency and being part of the PAP machinery, because Dr. Fang has set the bar so low, Mr. Yip comes across as comparatively prepared and polished. His PAP profile page describes his teaching and civil service background (also, his 2020 introductory video largely answers the three questions Prof. Teo posed), and he has his own website alongside his Facebook and Instagram pages. Substantively too, we often hear about Louis Ng as the most active MP in the 14th term of parliament, but guess who’s number two (across the board, not just among PAP MPs)? It’s Mr. Yip.

What works with new candidate introductions?

Nevertheless, not all contests will be so cut-and-dried. Other candidates and their political parties are likely to have more decent online and in-person presence. Moreover, new candidates cannot draw from a parliamentary track record. As such, in the past weeks, having read, watched, and listened, I think new candidate introductions can be rendered less banal if I get to learn much more (i.e., more quantitative), in greater detail (i.e., more qualitative), and vis-à-vis their legislative aspirations (i.e., more substantive).

I want more quantitatively. Thus far, the PAP and the main opposition parties (i.e., WP, PSP, and PAP) have primarily relied on their press conferences covered by the mainstream media. Overall, in the lead up to Nomination Day, the WP seems to be presenting the most detailed information about their candidates, with the website noting their “Vision for Singapore” and “Philosophy and Passions.” I especially like the four- to seven-minute video introductions they’ve produced for each candidate, which are also completely centred on the candidates and their motivations.

The PSP has published profiles and videos of its new candidates, though they are a little briefer. At the moment, the SDP has a video featuring its potential and new candidates, but its website offers no additional candidate information. Conversely, the PAP publishes its candidate profiles as they are announced, though the write-ups are similarly sparse. Instead, the PAP may be making a big social media play, given that its candidates have the most populated Facebook and Instagram (and TikTok) pages.

I want more qualitatively. Yet, I am absolutely greedy to learn much more about the candidates in greater detail. For instance, with the PAP, I fully understand how their new candidates’ curated social media profile pages complement the party’s municipal strengths and door-to-door ground game during election campaigns. Strategically, however, these are approaches easily replicated by the other parties and candidates. It’s not particularly difficult – or meaningful – to exclusively post walkabout photographs and selfies or wefies, snippets of one’s hobbies and personal lives, or videos of your party volunteers’ tireless contributions or acts of kindness.

I get it. This may be a political era of short attention spans. Not everyone cares to know more about the candidates or their potential parliamentarians. And maybe senior minister Lee is right, that our voting decisions are guided more by the candidate’s party. At the same time, I don’t think it’s enough to let one’s party manifesto do all the heavy substantive lifting. In a very thoughtful manner, it’s not too much to ask a candidate to write about or speak to perceived challenges in Singapore society, while explaining the issues for which they’ll advocate.

And I want even more, substantively. Across both sides of the political aisle, requests for greater substance is neither outlandish nor difficult. For the new PAP candidates, beyond promises to serve residents in your neighbourhood, explicate how you intend to legislate and how you are qualified to legislate, likely as a backbencher.

For new opposition candidates, I agree that opposition for opposition’s sake is not adequate. Furthermore, besides staking your positions in relation to the ruling party, you are entering a context where your party colleagues have some legislative experience as MPs or NCMPs. If so, what have you learnt through observation or legislative assistance? What would future collaboration look like? And similarly, how are you qualified to legislate, likely in the minority?

Everything considered, it can feel like we’re asking for a lot from the candidates. Nonetheless, elections happen once every five years in Singapore, and the intense nine days of campaigning create a unique opportunity for voters to engage civically and politically. We owe ourselves that much. As Prof. Teo concluded, in the video explainer I referenced earlier:

“Elections and our voting decisions matter. We should think of our votes as having an impact on the kind of society we want to be. We should remember that we are not just voting for people to take care for municipal issues, but members of parliament.”

Last week, I wrote about the obvious prognostications – especially related to social service and social policy issues – for GE2025. Economically, inflation and the cost of living are likely top of mind. Experiences of poverty and inequality, or being part of the (perceived) “sandwich class,” will shape those inflation and cost of living concerns. And tied closely to economic concerns are immigration ones, alongside persistent expressions of xenophobia.

This week, I’d like to go beyond these expected issues to consider the social service and social policy topics I would like parties and candidates to discuss. They are broadly summarised as cohesion (of our communities), cultivation (of the young), caregiving (of the elderly), and compassion (for the disadvantaged). In doing so, I draw from recent commentaries and add particular questions of interest.

Because thus far, any meaningful contestation of ideas and policies has been overshadowed by the slow-drip of candidate-centred announcements. Questions posed to and features of these candidates largely portray them as milquetoast and likely to toe the party line. Yes, diversity and representation matter. Yes, growing up disadvantaged or marginalised allows for their lived experiences to inform policymaking. Yes, general interest in socio-economic and municipal issues is useful (perhaps expected). But, especially those representing the ruling People’s Action Party (PAP), I do not think it is too much to ask of potential parliamentarians – beyond what their parties offer – about their individual policy and legislative aspirations.

Cohesion (of our Communities)

The cohesion of our communities is linked to the government’s relationship to its people, especially given its outsized role across all aspects of our lives in Singapore. But how we relate to each other – as a population – matters too. This discussion was front-and-centre in January this year, at the annual Institute of Policy Studies conference. A key study finding that “the number of close friends that Singaporeans have has shrunk over the past six years, from 10.67 on average in 2018 to 6.49 in 2024” appears to suggest that our country’s social cohesion is strained. Additionally, Singaporeans appear to prefer interacting with others of similar income levels.

At the same conference, in his assessment of the health of community in Singapore, Hong Kong Baptist University’s Prof. Cherian George argued that “we don’t have the kind of horizontal, people-to-people trust that would allow community to flourish.” The solution is not for the PAP government to get out of the way or avoid “crowding out” diverse forms of civic engagement and political participation. Instead, cohesion is strengthened when the state is less autocratic and tolerates competition and contrarian perspectives in a more vibrant civil society. He concluded:

Also at the conference, Nanyang Technological University’s Prof. Teo You Yenn focused on the responsibilities of companies and corporations. “Occasional activities” such as corporate social responsibility, days of service, and volunteerism or philanthropy are not enough. She posed a list of questions to businesses, for example, in relation to wages and labour, whether workers have a choice in decisions affecting their well-being, and how businesses weigh their potential societal contribution and harm. “Too often, corporations exert unfettered power over lives because too small a group of people within them have had too much say in the decisions that govern large numbers of people.”

Here are some questions I’ll be looking out for:

1. How do political parties and their candidates understand “the community?” And how do they perceive growing socio-political pluralism in Singapore?

2. What do they think about activism and advocacy as they relate to issues (e.g., Gaza and Palestine, the death penalty, climate change and action) as well as engagement and participation forms (e.g., community organising, social media campaigns, university-based action)?

3. Who are the Singaporeans least likely to be engaged civically and politically – including and especially across past national conversations – and how do we reach them and better represent their voices?

Cultivation (of the Young)

A personal paradox clouds my assessment of Singapore’s education system. The system stressed me, it made me irrationally competitive about everything, and the mere thought of the high-stakes examinations I took decades in the past – and the worry I felt when waiting for my results – still gives me anxiety today. However, somewhat strangely, the wretched system and its unrelenting rigour gave me the knowledge and skills to succeed academically and professionally.

Still, most agree that academic pressures to do well are unhealthily high, for a Singaporean’s competitive rat race often begins in school (or even pre-school these days). My experience of benefits is also by no means representative. In this vein, nothing epitomises the challenges associated with the cultivation of Singapore’s students more than the reliable growth of private tuition. Our collective household expenditure on private tuition stood at S$1.8 billion in 2023, compared to S$1.4 billion in 2018 and S$1.1 billion. There is also an equity problem: “The top 20 per cent of households by income spent on average S$162.60 a month on tuition – more than four times the S$36.30 spent by the bottom 20 per cent.”

Serious criticisms of our education system are not new, though parents – previously educated in these pressure-cooker environments themselves, now witnessing their children going through the system and experiencing their pain by proxy, and perhaps still trapped in a rat race where comparisons vis-à-vis their peers are rife – have been increasingly vocal. Many find it hard to shake the appeal of perceived academic prestige and the impetus to have their children attend “elite schools,” particularly over neighbourhood ones. Relatedly, familiar policy recommendations include smaller class sizes, reforming the PSLE or making it optional, improved professional support for learning and mental health needs, as well as a more equitable and inclusive admissions system.

Here are some questions I’ll be looking out for:

1. Everyone has critiques of Singapore’s education system, yet there are objective measures of its relative success (e.g., PISA, TIMSS, comparative progress among low-progress learners). Hence, as a starting point, what has worked or been effective so far, especially since most political candidates are likely to have, to a pretty large extent, benefited from the system?

2. Thereafter, what needs to change? Which policies need to move together? The Ministry of Education is not oblivious to the aforementioned challenges and has taken a stab at reform. Political parties and candidates cannot be oblivious and propose changes in a vaccum. As such, how has the ministry done? What should it do?

3. What do we do about private tuition, given its outsized financial footprint and assumed necessity among parents and students? It will be hard to put the genie back in the bottle, but surely we could do more besides an advertisement code of conduct for the industry?

Caregiving (of the Elderly)

Caregiving of Singapore’s elderly “works” to some extent, for now, because we have collectively accepted three related realities. First, that family members – often the unmarried, eldest daughter without children – will step in and up as the primary caregiver. Moreover, such familial obligation is expected, and the associated burdens go unquestioned. Second, that we will continue our collective reliance on low-wage migrant domestic workers (MDWs) from Singapore’s neighbouring countries, some of whom shoulder household chores and child-rearing duties on top of their caregiving responsibilities. And third, that with enough money, wealthier Singaporeans can pay for higher-quality formal care.

Public concern over caregiving gained traction immediately after this year’s Budget. A forum letter to “The Straits Times” advocated for caregivers to be valued and financially compensated: “Moving beyond traditional support measures to recognising caregiving as essential labour.” Cumulatively, caregivers’ labour is valued at S$1.28 billion annually. Underlying this advocacy was also a desire for caregiving to not remain an “individual responsibility,” because family members and MDWs “perform the lion’s share of care.” Other relevant policy recommendations include paid care leave as well as the raising of the current Home Caregiving Grant for non-working, full-time caregivers.

Here are some questions I’ll be looking out for:

1. What is the balance or relationship between one-off Budget giveaways and permanent financial and social assistance programmes? Tied to such a question, I believe, should involve a nation-wide conversation about what graceful and meaningful ageing look like.

2. Relatedly, what does a good life look like, for full-time informal caregivers in the family?

3. What might caregiving look like, if informal caregivers were better compensated and were better taken care of? What might caregiving look like, if and when some family members – previously obligated and even guilted – are physically exhausted and psychologically burnt out?

4. How might Singapore cope, if and when the supply of MDWs run dry?

Compassion (for the Disadvantaged)

Meritocracy remains central to Singapore’s narrative and expectations of its continued socio-economic success, that hard work and excellence – paired with the availability of diverse opportunities – should result in chances of success, “regardless of starting point or family background.” To the government’s credit, discussions of income inequality and social mobility are no longer taboo. Furthermore, recognising that some Singaporeans are disadvantaged by skewed starting points and growing up in less-privileged family backgrounds, it has offered more targeted support to lower-income families and their children.

Yet luck is an important determinant of success in one’s life in Singapore. For example, the socio-economic status to which one is born, or one’s academic aptitude in a Singaporean education system which still privileges grades and credentials, may influence whether one emerges as a meritocratic “winner” or “loser,” and thus solidifying one’s ossified station in life. Politically (as mentioned in the last newsletter), as objective markers of socio-economic progress, the ruling PAP will point to rising household incomes, lower income equality, and fewer households requiring financial assistance. However, the government has to tackle more subjective voter perspectives about fairness and deservedness, feelings of being left out or left behind, and perceived obligations – or lack thereof – to those who are disadvantaged or marginalised.

Consequently, Singaporeans who see themselves as having “earned” their success, meritocratically, may hold less compassion for those who may not have done as well, when they are seen as not having worked hard enough.

Finally, compassion must also extend to our regional neighbours. The Trump administration’s scaling back of foreign aid – especially by gutting USAID (US Agency for International Development) – has not significantly affected Singapore, but it has been deleterious for South East Asia. A February survey by the philanthropy platform Asia Philanthropy Circle estimated hundreds of millions of lost funding dollars as well as millions affected by programme closures, service cutbacks, and staff educations. Malaysian non-profit organisations and social service programmes have been downsized or shut down, and US humanitarian support in the aftermath of the Myanmar earthquake has been compromised.

Here are some questions I’ll be looking out for:

1. How will parties approach the centrality of meritocracy in Singapore, vis-à-vis ongoing discourse around income inequality and social mobility? How do candidates individually understand meritocracy and their personal relationship to it?

2. What are the structural and systemic discussions surrounding disadvantaged and marginalised Singaporeans?

3. Amidst ostensibly meritocratic competition, what obligations do we have to each other as Singaporeans?

4. Given broader geopolitical trends, especially the US’s reduced footprint in the region, what is Singapore’s role? Should it have one?

Global news roundup

In South Korea, politically, impeached president Yoon Suk Yeol was released after a ruling that he was detained illegally, but he still faces insurrection and other charges. Then, the impeachment of a former prime minister was overturned by a court, thereby reinstating him as acting president.

The Truth and Reconciliation Commission found that adoption agencies committed widespread malpractices decades ago, when children were sent to American and European homes. The report was the government’s first official admission of problems related to rights violations of South Korean adoptees. Separately, at least 15 people were wounded after two of its fighter jets accidentally dropped bombs on a civilian area during a live-fire exercise. And multiple wildfires, the worst in the country’s history, killed at least 24 people.

In Syria, the new government has been struggling to restore control, following clashes between state military forces and gunmen loyal to the ousted president Bashar al-Assad. The majority of victims also appear to be civilians targeted in revenge killings. With thousands dead and curfews imposed, it is the worst violence since Mr. Assad’s ousting in December 2024. Later, the Kurdish-led militia which controls part of the country agreed to merge with the new government, a breakthrough which signifies a major step towards unification.

And in Turkey, authorities detained the president's main political rival on charges of corruption and aiding a terrorist group, capping a months-long legal crackdown on opposition figures. It has resulted in the country’s largest protests in over a decade.

Also read about the following in this month’s roundup:

• Elections in the horizon and the present;

• Diplomatic and geopolitical conflicts; as well as

• Political consequences and repercussions.

Singapore’s social service sector has helped shape social policy discourse, and we should continue to shape such discourse for GE2025 and beyond. With the election very close in the horizon, during an intense period of contestation, I plan to review and critique the emerging discourse (e.g., party manifestoes, campaign speeches, broadcast debates, candidates’ policy perspectives, news stories), focusing specifically on social service and social policy issues.

Because more important than the election results per se – or the ensuing political discussions about the ruling party’s mandate and the role of the opposition, which I’ll leave to my political science colleagues – are the discursive issues shaping Singapore’s socio-economic future. Unironically, this is somewhat consistent with the 2022 Forward SG Exercise, which sought to understand “how we should refresh our social compact for the road ahead” (emphases mine), a message likely to be echoed in the People’s Action Party’s (PAP) GE2025 manifesto.

I start with a two-parter, populating a social service and social policy discursive wish-list for GE2025. In this first newsletter, I argue that our sector’s proximity and familiarity – to individuals, families, and communities on a day-to-day basis – position us well vis-à-vis social policy discourse. Then, besides stating the obvious about bread-and-butter issues and immigration dominating the election, I add particular questions of interest. Next week, in a second newsletter, I go beyond these expected issues to consider the social service and social policy topics I would like parties and candidates to discuss. They are broadly summarised as cohesion (of our communities), cultivation (of the young), caregiving (of the elderly), and compassion (for the disadvantaged).

Proximity and familiarity: Singapore’s social service sector and macro social work

Every day, social workers and social service professionals in Singapore journey with and advocate for individuals, families, and communities – often low-income, disadvantaged, or marginalised – to help meet their basic needs and improve their lives. Through these interactions, we gain proximity and familiarity. Proximity, by witnessing and understanding the socio-economic challenges experienced by those individuals, families, and communities. Familiarity, in terms of the environments and systems they have to navigate, or the policies that we navigate with or for them, to access assistance and resources.

Such proximity and familiarity place the social service sector in a unique position vis-à-vis social policy discourse, because we draw from lived experiences and observations of how we could, collectively, do better. However, engagement in such discourse – especially in the Singaporean context – feels uncomfortable for some. I understand the discomfort. Nevertheless, these tendencies run counter to a broader social work framework, by encompassing micro and mezzo work but excluding macro areas. We work with individuals and families, through casework, counselling, and therapy (i.e., micro work), further engender change across groups, organisations, and communities (i.e., mezzo work), yet we often steer clear of systemic advocacy and policy activism (i.e., macro work).

Still, perhaps somewhat anecdotally, the willingness to engage in macro advocacy and activism appears to be shifting, albeit gradually. Not as overt mobilisation or political contestation, not yet, but – at the very least – the social service sector is helping shape Singapore’s social policy discourse. Just with the socialservice.sg podcast alone, over the past five years, I’ve documented social service researchers and practitioners:

• Examining poverty and insecurity, across food insecurity (including during COVID-19), housing and income insecurity, and precarity among migrant workers;

• Discussing different forms of civic engagement and political participation, including and especially during GE2020 and PE2023; as well as

• Engaging with community development in social work practice and reflecting on the privileged position of researchers.

With GE2025 just around the corner, I’d like for us, as a sector, to continue shaping the discourse for the election and beyond. Discourse is not only shaped by political parties and their candidates. Singapore elections offer an intense period of civic engagement and political participation, during which we’ll have the public’s attention, thereby presenting a unique opportunity to discuss social service and social policy issues which matter to us.

Bread and butter with a slice of immigration

Following the report by the Electoral Boundaries Review Committee and the gradual introduction of potential candidates, GE2025 is very, very near. Already, the following prognostications – related to social service and social policy issues – appear pretty clear to me.

“It’s the economy, stupid?”

Inflation and the cost of living are likely top of mind. If you need empirical evidence, earlier this year, the 1,130 respondents in a study by market research company Blackbox Research’s sentiment tracker SensingSG ranked their top voting concerns as: (1) cost of living and inflation; (2) jobs and employment; and (3) the economy. Relatedly, housing costs and support for those on lower incomes ranked seventh and tenth respectively.

For prime minister Lawrence Wong and his PAP, the decision not to call for snap elections last year – an exceptionally bad year for incumbent parties globally – seems validated too, given that more Singaporeans, in the same SensingSG study, now believe that the country is headed in the right direction (88 per cent in 2025, compared to 77 per cent in 2024). Similarly, in 2025, 78% of respondents rated their financial situation to be “quite good” or better (compared to 64% in 2024).

In addition to Budget 2025 – characterised by “The Straits Times” as a “bonanza Budget” – the prime minister will contend, based on the numbers, that Singapore is on the right economic track. Mr. Wong and the PAP will point to rising household incomes, lower income equality (after accounting for taxes and government assistance), and fewer households requiring short-to-medium and long-term assistance (which I covered in a past newsletter). During election hustings, expect persistent references to ComLink+, a relatively young programme for lower-income families with children.

Relatedly, experiences of poverty and inequality – or being part of the (perceived) “sandwich class” – will shape those inflation and cost of living concerns, especially as they relate to housing, education, as well as employment and income concerns. Put otherwise: If there is strong data that Singapore’s economy is doing well (i.e., higher household incomes and lower income inequality, as aforementioned, as well as the country’s inflation growth hitting a four-year low), why might some Singaporeans still feel bad about the economy?

Some questions I’ll be looking out for:

1. How will the parties juxtapose the objective economic conditions (i.e., strong economic indicators) and subjective experiences about income inequality and social mobility? (It’s a tad esoteric, but I recently published a journal article about how subjective social mobility in Singapore is associated with personal and societal optimism as well as political opinions.)

2. What are the parties’ messages to low-income households and those in the “sandwich class,” vis-à-vis specific issues (e.g., housing, COE and food prices, shrinkflation)? (In a podcast episode with Prof. Linda Lim of the University of Michigan, we discussed other important economic indicators and the persistent challenge of low wages in Singapore.)

3. How will Singapore be affected by macro-economic conditions, such as the Trump administration’s recent tariffs and the threat of a global trade war?

Singapore, first? Issues of migration, citizenship, multiculturalism, and nationalism

Tied closely to economic concerns are immigration ones, alongside persistent expressions of xenophobia. Globally too, characterised by nativist leaders like United States president Donald Trump and the growth of right-wing populism in European Union countries, immigration issues have heavily influenced elections. In Singapore, I’m especially interested in how immigration intersects with class, for example, pairing advocacy for “true-blue Singaporeans” over “foreigners,” while relying on the low-wage labour of migrant workers and migrant domestic workers.

Immigration has featured prominently in our politics in the past decade. With Dr. Rebecca Tan of the National University of Singapore, in a podcast episode, we covered five key events: (1) GE2011; (2) The 2013 Population White Paper; (3), The 2013 Little India Riot; (4), The focus on the 10-million population figure during and after GE2020; and (5), COVID-19 and the plight of migrant workers.

First, I remember GE2011 as the PAP’s poorest electoral performance, based on its popular vote share since independence. That happened, in my recollection, when there was a lot of frustration over crowded and unreliable public transportation, high HDB and BTO prices, as well as continuing economic challenges in the aftermath of the 2008-09 Great Recession. But it also felt like the “foreigner,” the “foreign talent,” and the “migrant worker” – who made up about 36 per cent of Singapore’s total population then – were conveniently framed as the “bogeyman.” For instance, in Aljunied GRC, voter opposition against the construction of a dormitory in Serangoon Gardens was not insignificant.

The second episode involves the 2013 Population White Paper, which controversially projected Singapore's population to hit 6.9 million by 2030. In particular, through many letters to the editor and in-person public protests at Hong Lim Park, I remember the public anger against immigration and our reliance on migrant labour. The third event is the 2013 Little India Riot, when general population overcrowding and specific congregations of migrant workers in public spaces were highlighted. The fourth set of events centre on the 10-million population figure, as a population figure, planning parameter, or target (and a cudgel against the government). Finally, during and beyond the COVID-19 circuit-breakers, there was a temporary increase in sympathy for the migrant workers stuck in the dormitories. Yet, it feels like both public attention and concern have since waned.

Some questions I’ll be looking out for:

1. How will the parties talk about immigration, especially against recent history in Singapore and recent global developments?

2. When evaluating employment statistics between Singapore citizens, permanent residents, and foreigners, what exactly are the policy prescriptions, while considering the country’s long-term demographics?

3. Beyond the economic and employment dimensions, how do immigrants fit with Singapore’s social fabric?

4. What will the parties say about or advocate for migrant workers and migrant domestic workers, who do not have a vote, but relied upon so heavily by Singaporeans?

“Coalitions of electoral losers”

“Coalitions of electoral losers” was how author Loke Hoe-Yeong – in a podcast episode – described Singapore’s opposition parties which are not the Workers’ Party (WP), the Progress Singapore Party, or the Singapore Democratic Party. Already, the WP has been clear and consistent with the goal of having elected opposition MPs make up a third of parliament. Claims of gerrymandering notwithstanding, with ongoing negotiations, everyone recognises that multi-cornered fights will adversely affect the opposition’s prospects.

On the other hand, for the ruling PAP, we foresee three scenarios, all of which we posed in a podcast episode to Dr. Kenneth Paul Tan of Hong Kong Baptist University. First, if the PAP scores five to 10 percentage points higher than they did in 2020; second, if they do about the same, losing a small handful of seats, but still maintaining their supermajority in parliament; and third, if the PAP is close to losing or even loses its parliamentary supermajority.

Elections and the (usually) nine-day campaign have always been exciting affairs. With a new prime minister at a helm and so much uncertainty in the air, GE2025 now feels qualitatively different.

And the issues which won’t – and shouldn’t – matter

Last but not least, the following issues are unlikely to matter: (a) altercations between party volunteers or spats during walkabouts; (b) the use of dialects in parliament; and (c) former nominated MPs resigning ahead of GE2025. They may, collectively, influence voter perceptions of the political parties – primarily the PAP – but I don’t think they warrant additional attention.

In a couple of days, in a second newsletter, I consider the social service and social policy topics that should matter. As aforementioned, they are broadly summarised as cohesion (of our communities), cultivation (of the young), caregiving (of the elderly), and compassion (for the disadvantaged).

Global news roundup (January and February)

January 2025

The increase in global temperatures reached 1.5 degrees Celsius – a threshold set in 2015 at the Paris Agreement – for the first time in 2024, also marking Earth’s warmest year on record since pre-industrial times. This was announced as, in the United States (US), Donald Trump was sworn in again as president, after he was voted out, impeached, indicted, and convicted of felonies. Mr. Trump is also pulling the US out of the Paris Agreement.

Also read about the following in this month’s roundup:

• Political turmoil in South Korea…;

• … And political turmoil around the world;

• Geopolitical tensions and other national tragedies; as well as

• Some good news from Asia.

February 2025

Ukraine marked three years of Russia's full-scale invasion. However, its geopolitical ties – especially with the United States (US) – have shifted dramatically. Following a call between US president Donald Trump and Russian president Vladimir Putin, the US met with their Russian counterparts in Saudi Arabia to discuss a path towards ending the three-year war in Ukraine. Ukraine and European countries were not included in the talks.

Furthermore, in escalating developments, US officials called on Europe to assume more control over its military and security situation, warned against Ukraine’s NATO membership, and said Ukraine may need to concede territory. In the United Nations, the US also opposed a resolution condemning Russian aggression in Ukraine. All these culminated in an explosive public confrontation which was televised, in which Mr. Trump and US vice president JD Vance berated Ukrainian leader Volodymyr Zelensky. The visit, which was meant to coordinate a peace plan and plans for a rare minerals deal, was cut short.

Overall, the moves signify a significant shift in US foreign policy under Mr. Trump, including closer links with Moscow.

Also read about the following in this month’s roundup:

• European politics;

• Other political and geopolitical developments; as well as

• Tragedy in Sweden and the US.

“Before the Ballot” podcast mini-series

Even though we probably jumped the gun, Singapore’s next general election must be held by November 2025. As part of our podcast mini-series this year, we interviewed eight guests across 11 episodes, spanning the economy, opposition parties, as well as migration, citizenship, and nationalism.

The socialservice.sg podcast is hosted on Spotify for Podcasters and is available on Apple Podcasts, Spotify, and on most streaming platforms. You can also follow us on Instagram (@socialservicesg).

1. GE2024? GE2025? "Before the Ballot" preview

2. “It’s the economy, stupid?”: Inflation, incomes, inequality, and Singapore’s general elections (with Dr. Linda Lim)

3. Singapore's opposition parties: The WP, PSP, SDP, and coalitions of electoral losers (with author Loke Hoe-Yeong)

4. Looking back and looking ahead: Dr. Kenneth Paul Tan on PAP’s governing principles, PM Lawrence Wong, and the next general election*

5. “Are we OK with the price of not being OK?”: The mental health and well-being of Singaporeans (with former NMP Anthea Ong)

6. Singapore, first? Issues of migration, citizenship, multiculturalism, and nationalism in Singapore (with Dr. Rebecca Tan)*

7. Parliamentary vacancies and parliamentary representation (with PhD student Joel Yew)

8. “Are unelected voices still necessary in parliament?” (with former NMP Anthea Ong)

9. Climate and environmental issues as election issues (with Bertrand Seah)

* Two-part episodes.

A voter sentiment survey: On swing voters and politicians’ personal lives

Digital magazine “Jom” commissioned its first-ever voter sentiment survey, involving 1,000 respondents. In the first part (published October 18, 2024), they examined political interests and imperatives, including Singapore’s most pressing political issues, Singaporeans’ personal financial situation, and general interest in politics and government. In the second (published October 25, 2024), respondents were quizzed about the country’s most popular politicians. Both articles are outside the paywall.

The two most interesting findings centred on the proportion of swing voters as well as whether improper conduct in a politician's personal life affected their suitability for office. In terms of swing voters, only approximately 26 per cent of respondents were “Very certain” about their party choice at the next election, and older Singaporeans were more likely to be confident of their vote. Notwithstanding the survey’s methodological limitations:

Political analysts generally assume that the swing voter segment in Singapore is around 35-40 percent of the electorate, bigger than in most democracies. Among other assumptions that feed this: the 35 odd percent who voted for Tony Tan, the PAP candidate, in the four-way 2011 presidential election are assumed to be PAP diehards; and the 30 odd percent that did not vote for Tharman Shanmugaratnam, the PAP candidate, in last year’s three-way presidential election are assumed to be ‘anything but PAP’.

Findings about the overall popularity and likeability of Singapore’s politicians were less interesting. However, it would appear that Singaporeans do generally consider politicians’ personal lives vis-à-vis their political suitability. About 29 per cent said it “Affects greatly,” 51 per cent said “Somewhat affects,” 15 per cent said “Doesn’t really affect,” and six per cent said “ Doesn’t affect at all.” Analytically:

“A conservative might be thrilled that almost 80 per cent of respondents feel a politician’s personal life (e.g. extramarital affairs) affects their suitability for office. A liberal, meanwhile, could reasonably argue that only less than a third believe it “affects greatly”. Put another way, over 70 percent of respondents may not have strongly objected if the philandering politicians were still in Parliament.”

Global news roundup

South Korea experienced 11 days of political turmoil. President Yoon Suk Yeol imposed a short-lived emergency martial law, but after the country’s legislature (with 190 of its 300 members present) voted unanimously to end it, he lifted the order. He survived a first impeachment effort after the ruling party boycotted the vote, but a second vote later passed. As a result, Mr. Yoon is suspended from office as the constitutional court deliberates on the impeachment vote.

Syria fell in 11 days. Aleppo, the country’s second-largest city, was first seized by rebel forces led by a Turkey-backed group, classified by the United States as a terrorist organisation. Later, the Syrian army withdrew rapidly from the strategic cities of Hama and Homs. The swift advance surprised observers. Ultimately, president Bashar al-Assad fled Syria to Russia as rebels quickly stormed and established control over the capital city of Damascus.

Also read about the following in this month’s roundup:

• Socio-political developments; as well as

• Geopolitical tensions.

Today, in this first newsletter on Christmas Eve, a panoply of government and research reports offered a composite picture of the average socio-economic life of the median Singaporean as well as that of those who are disadvantaged. We summarise those reports and highlight the key and interesting findings.

And in the second newsletter, ahead of Singapore’s next general election – which must be held by November 2025 – we do a quick recap of our “Before the Ballot” podcast mini-series and briefly examine a voter sentiment survey published by digital magazine “Jom” (whom we interviewed in a podcast episode back in 2022).

The average socio-economic life of the median Singaporean…

Economically, on average, Singaporeans appear to be faring better in a post-COVID-19 environment. Median-wage earners experienced increased incomes after accounting for inflation (at 3.4 per cent in 2024, an improvement from the -2.2 per cent in 2023). Also in terms of labour: Singapore’s unemployment rate (as at June 2024) was 2.7 per cent for PMETs (professionals, managers, executives, and technicians) and 3.4 per cent for non-PMETs.

Additionally, employed residents worked fewer hours (41.6 hours a week on average in 2024, compared to 46.6 in 2010), fewer residents engaged regularly in platform work (67,600 in 2024, compared to 70,500 in 2023), and an ageing population has resulted in the decline of Singapore’s labour force participation rate, of residents aged 15 and over, for the third consecutive year.

Individually, the median nominal income in 2024 – before accounting for inflation – was $5,500. The average Singaporean household earned $15,473 and spent $5,931 every month. In terms of what households spent: “Housing, food and transport topped the list of household expenses in 2023. They collectively accounted for 63.2 per cent of monthly household expenditure in 2023, compared with 64 per cent in 2017/18.”

… And disadvantaged Singaporeans

Persons with disabilities at the workplace

Worryingly, positive attitudes towards persons with disabilities (PWDs) have gone down. In the workplace – as more PWDs have been employed (including in full-time employment – just over half of Singaporean respondents (50.6 per cent) in the 2023 “Public Attitudes Survey” held positive sentiments towards PWDs. This is a nine percentage point drop from 59.6 per cent in 2019. More generally, 68.9 per cent of Singaporeans held a positive attitude towards PWDs in 2023, but this was also a drop from 2019, when the figure was 76.8 per cent.

Having regular contact with PWDs – at least twice a year – appears to improve the attitudes Singaporeans may have, thereby underscoring the importance of public education and anti-discrimination efforts. “More efforts are needed to raise awareness, foster understanding, and promote acceptance of [PWDs] and the challenges they face in the workplace.” These findings were all included in the recent “Disability Trends Report,” published by the Ministry of Social and Family Development (MSF) and the National Council of Social Service (NCSS).

Another potential challenge? Caregivers. The report found that between 2022 and 2024, the quality of life of caregivers of PWDs decreased from 64.7 to 62.1 (ranging from 0-100, with higher scores reflecting a higher quality of life). This contrasts with the higher quality of life scores reported by PWDs themselves. Specifically, for caregivers:

The decrease in [quality of life] scores … were significantly larger among caregivers of PWDs with low family resilience, high caregiving burden, and low perceived social support, compared to those with moderate to high family resilience, low caregiving burden, and high perceived social support.

Lower-income households in Singapore

MSF published its “Supporting Lower-Income Households Trends Report,” which documents the government’s approach towards socio-economic support for lower-income households. These were the key insights:

• Overall household income trends: Households in the first and second deciles (the two lowest deciles) experienced real increases in their average monthly household income from work, and the increases were higher than that of the median Singaporean household.

• Community Care (ComCare): Fewer households were assisted through short-to-medium term (SMTA) and long-term assistance (LTA) – Singapore’s key social safety nets – and a smaller proportion returned to receiving temporary financial aid (i.e., SMTA). These trends, MSF concluded, resulted from “the post-COVID-19 economic recovery, labour market expansion in recent years, and support received from other government schemes.”
  Quantitatively, from 2019 to 2023, the number of SMTA households decreased from 27,989 to 22,960, and the number of LTA households decreased from 4,175 to 3,479.

• ComLink+: Under this new 2023 scheme – to facilitate social mobility and reduce income inequality – family coaches and volunteer befrienders are key actors, who work to help lower-income households attain self-reliance. Less than half of these households are considered to be “stable” at the moment. Additional support measures in the areas of pre-school education, stable employment, debt clearance, and savings for home ownership was also recently announced.

Overall, including through ComLink+, the importance of boosting social mobility has been emphasised:

“If those who are poorer do not see real ways to move up the socioeconomic ladder despite their best efforts, it will breed a sense of unfairness and the feeling that the system is stacked against them … This will lead them to view others with suspicion, and threaten the country’s ability to tackle important issues together.”

What also warrants further interrogation (which is especially exciting for the evaluators and practice researchers among us) are the “desired outcomes” MSF has identified for ComLink+ households to determine “stability.” The framework is loosely organised as a truncated logic model (above). As aforementioned, less than half of the households are considered “stable” right now, and these outcomes do invite further questions, about theories of change and trajectories of progress (i.e., How exactly do households move from stability to self-reliance to social mobility, and how exactly do the family coaches feature?) and how the outcomes are operationalised (e.g., How do we determine if households are “safe from harm” and that family relationships are “resilient”?)

Mental health stigma and youth mental health development

Even as awareness of mental health conditions has increased and stigma has decreased, the Institute of Mental Health (IMH) continues its work in research and public education, especially among young Singaporeans. In its second “Mind Matters: A Study of Mental Health Literacy” report, involving 4,195 respondents, 63.3 per cent correctly recognised common mental health conditions.

Relatedly, stigma scores – measured by “the extent to which respondents viewed those having mental health conditions as dangerous or unpredictable and if that was a sign of personal weakness rather than a medical condition, and if respondents wanted to socially distance themselves” – decreased, thus reflecting improved public perception of individuals with mental health conditions.

Separately, the IMH has launched a decade-long study “to support and track the development of youths with certain mental health and neuro-development conditions.”

Global news roundup

Dishearteningly, a United Nations report said that countries have made no progress in cutting emissions resulting from the use of oil, gas, and coal. No progress has also been made in tackling global warming. Later in the month, at the end of COP29, wealthy countries pledged to provide S$404 (US$300) billion annually to poorer countries so as to cope with the climate crisis and its impact. However, the figure has been criticised as woefully insufficient.

These happened as over 150 were killed in landslides and massive flooding in the Philippines, and historic flash flooding in Spain - the country's worst in at least three decades – left over 200 dead. Angry protestors shouted and threw mud at the king and prime minister.

In Israel, the defence minister was fired. Then, the International Criminal Court issued arrest warrants for the Israeli prime minister and the former defence minister, for crimes against humanity and war crimes. A warrant was also issued for Hamas’s military chief, who is said to have been killed in an Israeli airstrike. Finally, the security cabinet voted in favour of a ceasefire deal to end the fighting with militant group Hezbollah in Lebanon.

And Ukraine marked the 1,000th day of the deadliest conflict in European history since the Second World War. Russia struck Ukraine with intermediate-range missiles, as Ukraine struck Russia with United States (US)-supplied short-range ballistic missiles. The US also confirmed reports of North Korean troops operating in Russia.

Also read about the following in this month’s roundup:

• Trump’s triumph and other political developments;

• Strained geopolitics; as well as

• Legislative contentions.

With caregiving and caregiving support, a Duke-NUS study estimated informal caregiving time and its annual monetary value, while an Institute of Mental Health study examined caregivers of individuals with dementia. A Lien Foundation and Kantar survey studied Singaporeans’ awareness and knowledge of palliative care, and another report underlined preferences of older adults to age in place. And with youth mental health, given that the struggles of those who are depressed, anxious, and stressed are well-understood, what are we actually going to do about the problems?

Finally, I share percentages or proportions related to financial freedom, workplace discrimination, and perceptions of democracy. And, as always, we conclude with your monthly global news roundup (September and October).

Spotlights on caregiving, palliative care, and youth mental health

Caregiving support and support to caregivers

Caregivers in Singapore provide important support – practically and monetarily – yet support to caregivers warrants further attention. A Duke-NUS study estimated that caregivers provided an average of 33.0 hours of weekly care, and seniors, in turn, received 60.5 hours of care from their primary and secondary caregivers. Cumulatively, the caregivers’ labour was valued at S$1.28 billion annually.

However, it is noteworthy that migrant domestic workers (MDWs) – as live-in full-time employees – provided even more caregiving labour (42.0 hours a week) than the caregivers, reflecting our persistent reliance on them. Their labour is especially significant, considering the other household duties for which they are responsible as well as their meagre compensation. In fact, during his visit to Singapore, Pope Francis called on our political leaders to seek fair wages for migrant workers (including the 290,000 MDWs and 440,000 workers in the construction, marine, and process sectors).

Caregivers, including the MDWs, will also have dementia in their sights. Although dementia prevalence among older adults aged 60 and above – resulting from higher employment and education and lower stroke prevalence – has decreased from 10 to 8.8 per cent between 2013 and 2023, the absolute number of Singaporeans with dementia will continue to grow. Consequently, given the increased demand for care needs, care burden and psychological distress among caregivers are expected to rise.

More precisely, this second Well-being of the Singapore Elderly (Wise) study by the Institute of Mental Health (IMH), involving 2,010 respondents, also found that caregivers of individuals with dementia reported significantly higher care burden and psychological distress, compared to caregivers who care for those without dementia.

Palliative care and end-of-life concerns

Palliative care seeks to improve the quality of life for those facing life-threatening illnesses, and the family of these individuals. Among Singaporeans aware of palliative care, while most understood its goals to maximise one’s quality of life and help the whole family cope with a serious illness, “many still mistake it as primarily care of the dying.” The Pulse on Palliative Care survey involving 1,005 Singaporeans – commissioned by family foundation Lien Foundation and conducted by market research company Kantar – was also conducted in 11 other Asia-Pacific regions or countries. Other critical themes included attitudes towards and experience with palliative care (evaluated by respondents to be mostly positive and accessible), as well as priorities and concerns towards death (with a majority wishing to be cared for and to die at home).

Similarly, a study by Ren Ci Hospital and Forbes Research with 500 respondents echoed these findings, concluding that palliative care knowledge appears limited to those who are already in the system and using healthcare services. Ren Ci clinical director David Ng said this to “The Straits Times”, that the aforementioned Pulse on Palliative Care survey:

“… Validated that many in the general population do not understand the full scope of palliative care, overlooking its broader scope to relieve symptoms and enhance quality of life throughout an illness trajectory.”

At the very end of their lives, approximately eight in 10 older adults (aged 53 to 79) prefer to age in their current homes, ideally in communities which housed eldercare or active ageing centres, transportation services, medical facilities, and nearby shops and eateries. This focus on ageing-in-place was included in a study of 6,020 respondents by the Centre for Research on Successful Ageing.

What’s next for youth mental health strategies?

So much has been made of the mental health struggles among young Singaporeans. Findings from the IMH’s first nationwide survey on youth mental health – primarily centred on high depression, anxiety, and stress prevalence – echo what we probably already know, yet discussions on policy changes or evidence-based practice implications feel scant. For instance, proposals to improve school mental health curriculum and help-seeking behaviours have been previously mooted, but our progress and evaluation of programme effectiveness are still unclear. Similarly, we’ve known about excessive social media use and cyberbullying’s ramifications too… What are we actually going to do about them?

Perhaps the mental health advocacy group, SG Mental Health Matters, comes closest. While its latest white paper is principally concerned with articulating a national suicide prevention strategy – which includes a national coordinating office, data-driven surveillance system, and national public awareness campaign – its systematic approach to implementing solutions is refreshing. Because, fundamentally, we want to understand the reasons for young Singaporeans experiencing depression, anxiety, and stress (as well as suicide ideation), such that efforts remedy causes in the earlier stages, and that interventions are not too piecemeal and do not come too late.

Socio-political percentages and proportions

44 per cent of Singaporeans said financial freedom is out of reach. The same proportion (44 per cent) also believed that having children will slow them down in being financially free, according to composite insurer Singlife’s second financial freedom index, which surveyed 3,000 individuals online.

Approximately 33 per cent of young Singaporeans have witnessed or experienced workplace discrimination, according to a poll of 1,006 18- to 43-year-olds conducted by market research company Kantar. This is despite a majority observing their workplaces to be inclusive, where discrimination is ostensibly minimised and diversity embraced.

72.2 per cent of Singaporeans strongly agreed or agreed with the statement, “Democracy may have its problems, but it is still the best form of government.” This was documented from the Singapore edition of the Asian Barometer survey, involving 1,000 participants aged 21 and older. Researchers at the Institute of Policy Studies also classified respondents into three archetypes “based on their contentment with democracy and the current system of government,” including those most content (22.1 per cent), in the middle (53.9 per cent), and those “ambivalent” (24.1 per cent).

Global news roundup

Across Lebanon, hundreds of coordinated pager explosions, for which Israel is said to be involved, killed dozens and injured many more. Reports alleged that Israel rigged the pagers with tiny explosive material and a switch before they were delivered to Lebanon. Days later, hundreds of walkie-talkies and solar devices detonated in a second round of explosions. Hezbollah vowed retribution, as the exploding devices marked a significant escalation in 11 months of conflict.

This was followed by hundreds of Israeli airstrikes against alleged Hezbollah targets – in the area’s deadliest day of attacks since Israel’s war with Hezbollah militants in 2006 – and culminated in the assassination of Hezbollah leader Sayyed Hassan Nasrallah. In response, Iran fired about 180 ballistic missiles at Israel. Weeks later, Hamas leader Yahya Sinwar, also the architect of the October 7 attack on Israel, was identified and killed in Gaza.

Also read about the following in this month’s roundup:

• Global geopolitical conflicts;

• Laws and judicial conflict; as well as

• Electoral and political developments.

In the first section, I take a deeper dive into (and critique) AWWA’s “Family Empowerment Programme”, which used a randomised controlled trial to study the impact of unconditional cash transfers in Singapore. Interestingly and recently, in the United States, four large-scale studies evaluated variations of unconditional cash transfers and a universal basic income, and I’ve included my thoughts in this newsletter too.

Next, we learn more about the Household Needs Study. Its approach to identify what items and activities were deemed “essential” to Singaporeans and whether we could access and afford these items, I believe, complements the past Minimum Income Standard studies. However, its methodological assessment of attitudes and beliefs regarding poverty was problematic. In this section, we also share a seven-part manifesto for a more socially and economically just country and review some unsurprising demographic and social trends related to Singaporean families.

Finally, I present percentages or proportions related to problematic smartphone use, wealth inequality, critical speech, as well as drug use and abuse. And, as always, we conclude with your monthly global news roundup (July and August).

Unconditional cash transfers and randomised controlled trials in Singapore and the US

A first in Singapore: AWWA’s “Family Empowerment Programme”

AWWA’s exploratory study involving low-income families received significant attention because its approach is unprecedented in Singapore. Unconditional cash transfers (i.e., money given without requirements and pre-conditions; though, in this study, participants had to qualify as a chronically low-income household) are rare in our social service sector, where means testing (i.e., cash transfers are conditional) is the norm. Additionally, AWWA evaluated the impact of these transfers using a randomised controlled trial (RCT), which is equally rare. We often describe the RCT as “the gold standard” in evaluation research, because the RCT design allows us to make causal claims more confidently.

Put otherwise: If you recall the common refrain, “Correlation is not causation,” the RCT gets us closer to causation. Causation about what? In this case, AWWA wanted to see if receiving unconditional cash transfers (S$500 monthly average, over 18 months) would cause significant improvements to education, skills upgrading, and employment (ESE).

Methodologically, a few notes about the study and evaluation design:

1. This is not a study of universal basic income (UBI). A UBI is universal (i.e., everyone gets it). Interestingly, a few US studies evaluating variations of the UBI or unconditional cash transfers were recently published, and I’ll share them later.

2. AWWA’s study is not a strict comparison between unconditional and conditional cash transfers, but one which compares: (a) receipt of unconditional and conditional, existing transfers; versus (b) receipt of the conditional, existing transfers only. In other words, the families were receiving these new unconditional cash transfers on top of their existing financial assistance and other social services.

3. I describe AWWA’s study as exploratory, which makes it a missed opportunity that the government has, at the moment, chosen not to replicate or scale it. More on this later too.

So, what did the study find? Substantively (and I want to stress that my critiques take nothing away from AWWA’s contributions), it is my opinion that the overall quantitative, RCT-based results were underwhelming.

1. Of the 17 indicators (12 interim and five final outcomes), only two were statistically significant in the hypothesised direction. In fact, none of the five final outcome indicators registered a statistically significant difference.

2. The good news, in practical terms, was that those who received the additional cash transfer (i.e., 75 in the “treatment” group) – compared to those who did not (i.e., 95 in the “control” group) – were more likely to obtain more secure employment contract types and held out for higher quality jobs without immediate financial stress.

3. Conversely, the not-so-good news was that those who received the cash transfer were also less motivated to pursue ESE opportunities and submitted fewer job applications.

When you read the report in detail, some indicators are reported as being statistically significant a little too loosely. If a key contribution of the AWWA study is its RCT design (i.e., we want to get to causation), to me, statistically significant differences between treatment (i.e., those who received the additional cash transfer) and control (i.e., those who did not) at the same time point is less notable than statistically significant differences between treatment and control over the course of the study (i.e., what the report itself describes as the difference-in-difference method).

(I will also add that the independent evaluator set the alpha statistical significance at a liberal 10%. Notwithstanding the sample size, if we used a more conservative and conventionally accepted level of 5%, none of the aforementioned indicators will be statistically significant. It would mean that, on all 17 indicators, we cannot conclude that there was a difference between receiving the additional cash transfer or not.)

What’s next in Singapore?

With the AWWA study, specifically, there are a few study and evaluation design tweaks I would consider:

1. Examine the first-order outcomes of the unconditional cash transfer itself: I personally think the 17 ESE indicators are second-order outcomes, preceded by – as the AWWA report itself documents – participants not having to worry about whether they qualify for assistance throughout the 18 months. That participants don’t have to worry about the constant flow of cash transfer is also a recurring theme from the qualitative data (i.e., the support does not have to be renewed).

2. Mediation analyses: AWWA could analyse not just how an indicator itself has changed over time, but how different indicators relate to one another. For example, improved mental health (at six months) could better job security (at 12 months), and result in higher income and more hours worked (at 18 months). Such evidence-supported narratives are currently missing.

3. Furthermore, AWWA could further exploit its longitudinal RCT design and analyse the other indicators related to family relationships and child outcomes. For instance, guided by the family stress model, we could test if reduced financial worry or anxiety improves parenting practices, and then enhances child development or flourishing.

4. What about the monthly cash assistance amount? It’s clear from the report that AWWA put in a lot of thought when deciding the actual cash amount, and I’m wondering if varying the monthly amount may make a difference on the ESE indicators. This question is of interest, when we read about a US study later.

In parliament, for now, Minister for Social and Family Development (MSF) Masagos Zulkifli said the government does not intend to replicate or scale the AWWA pilot, which I think is a shame. The minister is not wrong – as I’ve indicated above – about the study’s limitations. However, I believe we should be doing more of what AWWA has done, while addressing the shortcomings. Alongside the government’s ongoing evaluation of ComLink+ (including the recent inauguration of 200 family coaches and an initiative to encourage pre-school enrolment), what we collectively need are additional studies, research, and evidence in the Singaporean context.

Drawing from some US studies

Very recently too, four reports about guaranteed income in the US were published. The studies were all RCTs, though most also had a qualitative component (e.g., interviews or focus groups).

In Denver (Colorado), unconditional cash transfers were given to unhoused individuals, 631 of whom were studied. There were three payment groups. Groups A and B received a total of US$12,000 (S$15,800) over a year, but those in A received US$1,000 every month, whereas those in B received US$6,500 in the first month, then US$500 every month thereafter. Those in Group C, as the control group, received US$50 a month, for a total of US$600 a year (S$790). Notwithstanding the study limitations and the fact that participants were homeless to begin with, a key finding was that there were minimal differences in the targeted outcomes (e.g., housing security, financial well-being, and health) across Groups A, B, and C (i.e., the payment amounts did not seem to matter). As the researchers wrote:

“This may suggest that consistent cash assistance of even US$50 can improve many aspects of one’s life.”

In Los Angeles (California), 3,202 households (most of whom were in deep poverty) were given monthly unconditional cash payments (or guaranteed income) of US$1,000 (S$1,318) for 12 months. Significantly, those who received the payment reported reduced severity and frequency of intimate partner violence, were more likely to maintain their children's extracurricular activities (e.g., sports, after-school lessons), interacted more positively with their neighbours, and more likely to secure full-time employment. In this study, the theory of change was clearly fleshed out and tested: that the payments should reduce material hardship (or experiences of chronic scarcity), alleviate psychological stress (and improve cognitive capacity), and create “space for alternative ways of being and decision-making.”

A larger scale, longer-term project with two papers, likely the one referenced by minister Masagos Zulkifli, was implemented in Illinois and Texas. The 1,000 low-income participants (i.e., treatment group) received US$1,000 (S$1,318) per month for three years, and the control group received US$50 (S$65) per month. The first paper studied employment effects, and the second evaluated household consumption patterns.

With employment: Participants were more interested in entrepreneurship and were more willing to take risks. However, the transfer resulted in less individual and household earnings. Participants and their household members reduced their labour hours and spent more time on leisure activities, and there was no evidence for improvements to job quality or human capital. “For a policymaker interested in cash transfers, the main benefits would flow through the increased choice they offer participants in how to spend their time and invest for the future, even if relatively few use the opportunity for any one given pursuit such as as obtaining a post-secondary degree or starting a business.”

With household consumption patterns: Initial improvements to financial health did not last. “Household financial savings rise only modestly and financial assets are unchanged, while household debt rises slightly.” Additionally:

“Time-limited cash transfers may be a tool to increase spending on basic needs, reduce consumption volatility, improve financial health, and strengthen households’ ability to weather financial shocks in the short term, but policymakers interested in increasing wealth or improving long-run financial outcomes may want to consider either more permanent transfer programs or other policy options.”

Household needs, minimum income standards, and family trends

The Household Needs Study and poverty attitudes and beliefs

Two local researchers published a Household Needs Study (HNS), involving a two-step process. First, they identified what items and activities were deemed to be “essential” to Singaporeans (40 of the 51 items and activities were identified as essential, when at least half of the 4,014 respondents agreed). Second, based on the 40 essential items and activities, they estimated the proportion of respondents who could not access and afford them (about 30 per cent of the respondents were deprived of an average of four items or activities). Relatedly, they also found that most Singaporeans agree that “poverty is due to personal circumstances and actions, such as spending money on inappropriate items.” Few blamed bad luck or religious explanations.

The study complements and does not contradict the Minimum Income Standard (MIS) study (listen to our 2021 episode with lead authors Drs. Ng Kok Hoe and Teo You Yenn). The HNS focuses on the items and activities – as well as their accessibility and affordability – individually. Conversely, the MIS considers expenditure on goods and services at the household level and goes beyond what is “essential” to consider “a basic standard of living.” In other words, it is useful to know the individual items and activities that Singaporeans deem to be “essential” (i.e., the HNS approach), but we’d also like to understand how families approach basic needs vis-à-vis their budgets (i.e., the MIS approach).

More technically, I would characterise HNS as taking a variable-centred approach, and MIS as a more person-centred approach.

I do, however, take issue with how the recent HNS evaluated attitudes and beliefs regarding poverty. This is the evidence the researchers used to conclude that “within Singapore society attribution to poverty prioritises individual actions and circumstances”:

It could still be the case that most Singaporeans believe that low-income individuals have themselves to blame for being mired in poverty. Yet, there are methodological questions. First, how were the statements designed or decided in the first place? The researchers provided no such information. Second, relatedly, did the researchers have a priori hypotheses about how the statements may be grouped together, or were the categories decided later (i.e., individual actions and circumstances, wealth distribution in society, fate and supernatural forces)?

Third, given that respondents could agree with a statement and disagree with another, besides presenting the proportions as discrete or independent of one another, the researchers should’ve explored how respondents answered the questions collectively. For example, a respondent may agree that people are poor because they lack the ability to manage money, while concurring that there is an uneven distribution of wealth and that they have bad luck.

“The People’s Manifesto”

Ahead of Singapore’s next general election, a group of Singaporeans published a wide-ranging and detailed seven-part manifesto detailing proposed reforms which “are the most urgent for a more socially and economically just Singapore.” Some of the demands include (which might be most relevant to social service practitioners):

1. “To (re)build a democratic society”: Lowering the voting age to 18.

2. “Arresting the rising cost of living”: Expanding child and dependent care services.

3. “An economy that works for workers”: Implementing a universal minimum wage.

4. “Housing: Shelter, not commodity”: Improving the conditions of rental housing.

5. “Ensuring retirement adequacy”: Recognising a minimum standard of living.

6. “Equitable healthcare for all”: Introducing regulations to ensure safe working hours and staff-to-patient ratios for healthcare workers.

7. “Tackling the climate crisis with environmental justice”: Strengthening protections against unsafe working conditions.

These demands are a fraction of what is laid out in the document, and interested readers should also review the references provided by the authors.

“Family Trends” and play in Singapore

MSF’s annual “Family Trends” report presented demographic and social trends which were generally unsurprising. Worryingly, the number of marriages has decreased, Singaporeans are getting married for the first time later (though, interestingly, those who married at 20 to 24 years had higher marital dissolution rates), they are becoming first-time parents at later ages, and Singapore’s total fertility rate reached the historic low of 0.97 in 2023. Against the background of an ageing population, the number of elderly Singaporeans living alone – who rely heavily on family ties – doubled from 2013 to 2023.

Encouragingly, more fathers took government-paid paternity leave (from 47 per cent to 53 per cent between 2016 and 2021), and early childhood education appears more accessible and affordable.

Relatedly, Playeum – a charity focused on using play and the arts to create equitable learning opportunities for Singaporean children – published a white paper “to catalyse conversation about how we could enable our children to have more playful everyday lives.” Using existing local research, the paper neatly summarises how our children’s mental and physical well-being are jeopardised (e.g., insufficient and poor sleep, excessive screen and TV time, sedentary lifestyles and obesity, prevalence of mental health disorders). It also highlights some structural trends which have resulted in these challenges, such as high-stakes schooling and our excessive reliance and expenditures on tuition.

The central recommendation to “encourage our children to engage in everyday, creative free play (indoors or outdoors)” – facilitated by supportive adults – is compelling. The authors also offer some suggestions (e.g., “Let your child see you engage in a hobby that brings you joy”), some of which, presumably, are done on a small-scale by Playeum. Yet, first, they lack a Singapore-specific framework which could inform programmes, and second, the suggestion do not quite address the aforementioned structural trends the authors reported.

Social percentages and proportions

30.2%: How pervasive is problematic smartphone use, or PSU, in Singapore? Using a nation-wide population survey of 6,509 participants, a group of Singaporean researchers estimated that 30.2 per cent are likely addicted to smartphones (e.g., “Missing planned work due to smartphone use”, “The people around me tell me that I use my smartphone too much”). Those aged 15-34, with lower levels of education, or separated, widowed, or divorced were more likely to report PSU. Having PSU is also associated with psychological distress (i.e., depression, anxiety, and insomnia).

22.9%: The percentage increase in Singapore wealth’s inequality between 2008 and 2023, according to the UBS Global Wealth Report 2024. Specifically, “it appears that most of the rise in wealth has benefited the upper income brackets.”

Specifically, the report compared mean and median wealth. Singapore’s mean or average wealth increased by 116 per cent, but our median wealth decreased by two per cent. When mean wealth exceeds median wealth, that means those in higher wealth brackets experienced faster growth than those in the lower wealth brackets.

41%: The proportion of Singaporeans who say that people should not be able to publicly criticise the government. In this 2023 Pew Research Centre survey, a majority did say the opposite (i.e., 55 per cent of respondents agreed that Singaporeans should be able to criticise the government), but – of the nine East and South East Asian countries surveyed – Singaporean respondents were least supportive of critical speech.

18% of Singaporean youth know someone who has taken drugs. This is up from 10.6 per cent in 2019, based on surveys by the National Council Against Drug Abuse. The 2023 survey also found that, specifically, “among secondary school students, only 60 per cent of them agreed that drugs are more harmful than tobacco,” leading to concerns over more permissive attitudes towards drug and drug use.

Global news roundup

The Paris Olympics began with an extravagant opening ceremony, featuring nearly 7,000 athletes, and ended after two weeks of compelling storylines. Before the Olympics, in France, a coalition of leftist parties won the most seats in the final round of parliamentary elections, which came as an upset to the nationalist-populist National Rally, who had led the first round of voting. The voter turnout was the highest in four decades.

In Bangladesh, university students protested against a quota system which reserved 30 per cent of government jobs for relatives of veterans in the country’s 1971 independence war, and dozens have been killed following clashes with the police. A nationwide curfew was paired with an Internet shutdown. Later, as deadly anti-government protests continued and protestors stormed her official residence, the prime minister resigned and fled. Following her resignation, Nobel-winning microfinance pioneer Muhammad Yunus is set to lead a caretaker government.

Also read about the following in this month’s roundup:

• Joy in France and Bangladesh;

• UK and US electoral earthquakes;

• Geopolitical violence rages on; as well as

• Electoral and political developments.

On May 15, 2024, at the Istana, Singapore's fourth prime minister Lawrence Wong took his oath of office and was officially sworn in, taking over from prime minister Lee Hsien Loong. This political transition is significant for a host of different reasons, and so, across two episodes, we take a deeper dive with Dr. Kenneth Paul Tan of Hong Kong Baptist University. We look back, and we look ahead. We look back, across Singapore's first three prime ministers or three generations of leadership. We then look ahead to three potential electoral scenarios for Mr. Wong and his ruling People's Action Party or PAP.

In the first episode, we ask Dr. Tan: Across Lee Kuan Yew, Goh Chok Tong, and Lee Hsien Loong, how has the ruling PAP consistently approached Singapore’s governance since we became a self-governing state? We then explore those three electoral scenarios for Mr. Wong, in terms of the PAP doing better than they did in GE2020, the party doing about the same, or if they come close to losing or even loses its parliamentary supermajority. LISTEN TO THIS EPISODE.

In the second episode, we focus on the next general election. Fundamentally, we ask Dr. Tan, is it the PAP’s election to lose, or the opposition parties’ election to win? LISTEN TO THIS EPISODE.

We previously hosted Dr. Tan on this podcast in 2020, in an episode focused primarily on our collective, growing reliance on migrant labour in neo-authoritarian Singapore.

“Are we OK with the price of not being OK?”: The mental health and well-being of Singaporeans (with former NMP Anthea Ong)

The links between mental health and well-being and Singapore’s general elections may not be immediately clear. Yet, in GE2020, during a pandemic election which also placed emphasis on how Singaporeans were doing psychologically, many electoral candidates and political parties campaigned on related platforms. With former nominated member of parliament Anthea Ong, we first asked her to assess the current state of mental health and well-being in Singapore. Then, we discuss how these issues may feature in the next general election, after examining recent parliamentary amendments which were made and passed in the Law Enforcement and Other Matters Bill. LISTEN TO THIS EPISODE.

We previously hosted Anthea on this podcast in 2020 and 2021:

• Mental health: Awareness, acceptance, and action (A conversation with NMP Anthea Ong)

• “Are we listening?”: Coalescing around downstream and upstream mental health action (Anthea Ong; COVID-19, one year on)

• #AreWeOkay and the proposal for a Mental Wellbeing and Sustainable Development Office in Singapore

Global news roundup

Multilaterally, in the European Union (EU), centrist groups retained their majority in the EU parliament, though a few nationalist-populist countries made gains. In France, for instance, following his coalition's poor election results, president Emmanuel Macron called for snap parliamentary elections. Separately, the outgoing Dutch prime minister Mark Rutte was confirmed as the next secretary-general of the intergovernmental military alliance, NATO (North Atlantic Treaty Organisation).

Bilaterally, Russian president Vladimir Putin's first visit to North Korea, both countries signed a defence partnership to provide mutual aid if either is attacked. Later, the International Criminal Court issued arrest warrants for the country's defence minister and a general in the context of the situation in Ukraine.

Separately, in Yemen, Houthi rebels reportedly detained a group of international aid workers, including United Nations (UN) staffers.

Also read about the following in this month’s roundup:

• Israel’s Gaza invasion grinds on;

• Electoral victories, with asterisks; as well as

• In other news.