Key points at a glance:
- Recent activist surveys claim widespread discrimination against LGBTQ persons.
- They aim to change Singapore’s Workplace Discrimination law to enforce LGBTQ ideology in the workplace.
- Upon closer inspection, these surveys have severe methodological flaws that limit their reliability.
- The rise of academic activism threatens public trust in Singapore’s institutions.
A local study reported that 50% of LGBTQ respondents experienced workplace discrimination or harassment. It is a striking statistic and has been cited to expand legal protections for sexual orientation and gender identity (SOGI) under Singapore’s Workplace Fairness law.
If adopted uncritically, and lawmakers legally recognise SOGI because of this alleged “systemic discrimination”, the same justifications will then be used to advance policy change in education, healthcare, housing, family benefits and access to sex-specific facilities, which are often problematic themselves. We outlined these challenges here, here and here.
Given the potential for such broad legal changes, we need to examine carefully whether this study provides robust evidence for its claims.
The Researchers and Their Advocacy
The study in question is “Discrimination and Harassment in the Workplace” (Ng et al., 2023). Several authors have longstanding involvement in LGBTQ activism:
- Wen Zhi Ng – Hosted activities at Pink Dot through “The Courage Lab”.
- Rayner Kay Tan – Hosted activities at Pink Dot through “The Courage Lab”.
- Daryl Yang – Co-founded the Inter-University LGBT Network, launched at Pink Dot.
- Deryne Sim – Former Pink Dot Organising Committee member and spokesperson.
When researchers are closely associated with a protest movement to change Singapore’s laws, the study’s methodology deserves careful examination.
Importing American Academic Activism?
The survey was adapted from the Employment Experiences Survey developed by the Williams Institute at UCLA, where Deryne Sim completed her Master’s degree in Law and Sexuality. The Williams Institute was designed for legal and political LGBTQ activism, and directly influenced the US Supreme Court to legalise same-sex marriage (Obergefell v. Hodges).
Researchers regularly learn from overseas scholarship. However, we should distinguish between scholarship and academic activism. Scholarship seeks the truth even if it is politically inconvenient. But academic activism weaponises scholarship for political goals. Importing academic activism would reduce Singaporeans’ trust in science, and delegitimise our academic institutions.
How Reliable Is the 50% Claim?
To determine how authoritative this study is, we need to assess whether the study reliably estimates discrimination in Singapore’s LGBTQ population. Several methodological limitations deserve attention.
| Issue | Why It Matters |
| Self-selection bias | The researchers recruited participants through Pink Dot’s social media channels. Those with negative experiences may have been more likely to participate than those with neutral or positive experiences, making the sample less representative of Singapore’s broader LGBTQ population. |
| High attrition | Nearly 40% of respondents did not complete the survey. Without analysing those who dropped out, it is difficult to determine whether the final sample remained representative. It is possible that respondents dropped out because they felt that workplace discrimination is not part of their experience. |
| Limited verification | The anonymous online survey collected no personal identifiers to verify that respondents work in Singapore, or prevent multiple submissions for the purpose of inflating the numbers. |
| Undefined discrimination | The researchers did not provide respondents with a common definition of “discrimination”, allowing individuals to apply their own interpretations when reporting experiences. |
| Lifetime rather than current experiences | The survey measures lifetime experiences rather than recent incidents, making it impossible to determine whether workplace discrimination is increasing, decreasing, or remains stable over time. We don’t know whether these reported experiences are recent or from years ago. |
| Discrimination combined with harassment | Sexual harassment is already prohibited regardless of characteristics like sexual orientation. Combining harassment and discrimination into a single statistic is unnecessary and inflates the numbers when only discussing anti-discrimination legislation. |
| Self-reported experiences | The findings rely on respondents’ perceptions and cannot independently verify whether reported incidents meet an objective or legal standard of discrimination. |
Do Similar Surveys Confirm the Findings?
Researchers often try to prove the authority of their studies by showing that other studies give the same results. In this case, Pink Dot’s “No One Left Behind” survey and TransgenderSG’s “Challenges Facing Singapore’s Transgender Community” have reported similar findings and were cited as supporting evidence.
However, these studies have similar issues with their sampling methods and survey designs.
| Wen Zhi Ng’s “Discrimination and Harassment in the Workplace” | Pink Dot’s “No One Left Behind” | TransgenderSG’s “Challenges Facing Singapore’s Transgender Community” | |
| Advocacy Research | ⚠️ | ⚠️ | ⚠️ |
| Self-Selection Bias | ⚠️ | ⚠️ | ⚠️ |
| High Attrition | ⚠️ | ❓ | ❓ |
| Limited Verification of Respondents | ⚠️ | ❓ | ❓ |
| Undefined Meaning of “Discrimination” | ⚠️ | – | – |
| Overly Broad Meaning of “Discrimination” | – | ⚠️ | ⚠️ |
| Lifetime Rather Than Current Experiences | ⚠️ | ⚠️ | ⚠️ |
| Combining Discrimination and Harassment | ⚠️ | ⚠️ | ⚠️ |
| Self-Reported Experiences | ⚠️ | ⚠️ | ⚠️ |
That’s a lot of red flags for credible data to be reliably derived from. Mutually flawed surveys reasonably yield similarly flawed responses.
Overly Broad Meaning of “Discrimination”
A key difference between Ng et al. and later surveys by Pink Dot and TransgenderSG is how discrimination is defined.
While Ng et al. leaves the term undefined, the later two surveys give overly broad definitions of “discrimination” and “negative experiences.”
The surveys list the following statements as examples of discrimination. The researchers asked respondents to indicate whether they had experienced them using an Agree-Disagree scale (Likert scale). Based on these examples, the surveys construe mere non-affirmation of SOGI as discrimination. Needless to say, this exaggeration carries severe implications for freedom of speech and conscience.
| Survey Example | Possible Policy Implication |
| “Ability to bring your whole self to work (e.g. feel pressured to hide sexual orientation / gender identity or present a certain way)” | Employers may be forced to accommodate or affirm the open expression of LGBTQ identities. |
| “Being repeatedly and intentionally misgendered” | Employers may be forced to enforce preferred pronouns. |
| “Repeatedly addressed by my birth name, even when I’ve told them not to” | Employers may be forced to require the use of preferred names. |
| “Attempts to get me to change or reconsider my ‘lifestyle’” | Conversations about values may increasingly be treated as workplace misconduct or discrimination. |
| “Forced to use either the toilets for my assigned sex at birth or the unisex ones” | Employers may be forced to provide transgender persons access to opposite-sex toilets. |
Broadening the ordinary definition of discrimination naturally increases the number of experiences counted. This may help explain why other surveys reported higher rates than Ng et al. (2023):
- Ng et al. (2023): 50%
- Pink Dot (2024): 70%
- TransgenderSG (2021): 100%
If referring to a person by his or her biological sex is classified as discrimination, the policy implications extend beyond preventing unequal treatment. Regulators may expect employers to require the use of preferred pronouns, and treat disagreement as workplace misconduct.
Such a policy would compel employees to affirm contested beliefs about sex and gender, creating a direct conflict with freedom of speech and conscience.
From Research to Law
These consequences are not merely hypothetical. All three studies recommend policy changes that LGBTQ advocacy groups have pursued for years, including:
- recognising sexual orientation and gender identity (SOGI) as protected characteristics
- requiring employers to recognise preferred names and pronouns;
- regulating access to sex-specific facilities according to gender identity;
- extending employment-related benefits to same-sex couples; and
- introducing mandatory LGBTQ inclusion training.
This alignment does not by itself invalidate the studies. It does, however, make independence in research design especially important. When advocacy organisations define the categories, recruit the respondents and recommend policies based on the resulting findings, research easily becomes a vehicle for validating assumptions that were built into the study from the outset.
The United States illustrates the possible legal consequences. In 2024, the Equal Employment Opportunity Commission adopted guidance stating that repeated and intentional refusal to use a worker’s preferred name or pronouns, or denial of access to facilities corresponding to “gender identity”, could constitute unlawful workplace harassment.
The guidance, which has since been overturned, shows how an expansive definition of discrimination can be translated into rules governing speech and access to sex-specific facilities. Such rules may impede freedom of speech and conscience by compelling employees to participate in affirming contested ideological beliefs.
A Higher Evidential Standard
Lawmakers should hold research used to shape law to a high evidential standard, especially when its recommendations may affect speech, conscience and access to sex-specific facilities. Studies based on self-selected samples, subjective definitions and unverified responses may document the experiences of their participants, but they cannot by themselves establish the extent of systemic discrimination across Singapore.
This distinction matters. Advocacy groups are entitled to argue for legal change, but advocacy research should not be treated as neutral or representative evidence merely because it adopts the language and appearance of academic inquiry.
When methodological weaknesses are overlooked because the findings support a preferred policy outcome, public confidence in both research and policymaking is damaged. Singapore’s laws should therefore be shaped by evidence that is transparent, representative and able to withstand scrutiny, not by statistics designed around an advocacy agenda.
