Structural barriers to social protection and HIV prevention services for sex workers in Southeast Asia: a fixed-effects panel data analysis, 2018-2025

This fixed-effects panel data analysis of eight Southeast Asian countries from 2018 to 2025 reveals that while the immediate presence of structural barriers to social protection coincides with HIV prevention services for sex workers, their sustained existence significantly undermines the long-term availability and sustainability of these critical programs.

The Gist

The Big Picture: A Paradox of "Noticing" vs. "Fixing"

Imagine a country's health system as a garden. The "sex workers" in this story are a specific type of rare, delicate flower that needs special care to survive a dangerous pest (HIV). The "social protection barriers" are things like locked gates, thorny fences, or bad weather that make it hard for these flowers to get water and sunlight.

This study looked at 8 countries in Southeast Asia over 8 years (2018–2025) to answer a simple question: Does having these "locked gates" (barriers) help or hurt the delivery of special care (HIV prevention) for these flowers?

The answer is a bit like a plot twist in a movie: It depends on when you look.

---

1. The Immediate Effect: "The Alarm Clock"

What the data showed: In the exact same year that a country admits, "Hey, we have a locked gate blocking our flowers," they are more likely to have a special care team ready to help those flowers.

The Analogy: Think of this like a fire alarm.
When a fire alarm goes off (the country admits there is a barrier), the fire department (the HIV prevention program) immediately shows up with hoses and ladders. The alarm and the response happen at the same time.

• The Finding: The study found a strong link between reporting a problem and having a service ready. It seems that when governments finally say, "We have a problem," they often wake up and start building the solution immediately.

2. The Long-Term Effect: "The Rusty Gate"

What the data showed: When the researchers looked at the next year, the story changed completely. If the "locked gate" was still there the previous year, the special care team was less likely to be there the following year.

The Analogy: Now imagine the fire alarm went off, but the gate never got unlocked.
The fire department showed up once, but because the gate is still locked, they can't get back in easily next time. The garden gets neglected. The "locked gate" (the barrier) starts to rust and grow weeds, eventually making it impossible for the care team to do their job effectively.

• The Finding: The study found that if barriers persist, they act like a slow poison. They kill the momentum of the HIV programs. The initial "wake-up call" isn't enough; if you don't actually remove the barrier, the services disappear over time.

---

Why This Matters (The "So What?")

The researchers used a sophisticated statistical tool called a "Fixed-Effects Panel" (think of it as a time-lapse camera that watches the same countries over time to see what changes).

They discovered a paradox:

1. Short-term: Admitting you have a problem helps you get resources right now.
2. Long-term: Keeping the problem (the barrier) alive destroys those resources later.

The Takeaway for Policymakers:
It's not enough to just say, "We know there are barriers." That's like saying, "We know the gate is locked," and then walking away.

• The Mistake: Thinking that just having a program is enough.
• The Solution: You have to actually cut the lock off the gate. If you don't fix the structural barriers (the laws, the rules, the social stigma), the HIV prevention programs will eventually wither away, even if they started strong.

Summary in One Sentence

Noticing the problem gets you the fire truck today, but if you don't unlock the gate, the fire truck won't be able to come back tomorrow.

The authors are urging governments in Southeast Asia to stop just "reporting" the barriers and start aggressively removing them to ensure that life-saving HIV services don't just appear for a year and then vanish.

Technical Summary

1. Problem Statement and Research Gap

The global HIV response increasingly emphasizes addressing structural determinants (e.g., legal barriers, social protection access) to achieve epidemic control. However, there is a significant lack of robust empirical evidence linking specific national structural policies to the actual availability of essential HIV services for key populations in low- and middle-income settings.

While the conceptual relationship between policy adoption and health outcomes is understood, quantitative data on how specific structural policy indicators (such as legal barriers to social protection) correlate with the presence of HIV prevention services for sex workers in Southeast Asia is scarce. This study aims to fill this gap by quantifying the within-country relationships between national structural policy indicators and the availability of HIV prevention services targeting sex workers.

2. Methodology

Data Source and Sample:

• Source: Data was extracted from the UNAIDS Global AIDS Monitoring (GAM) framework, specifically the National Commitments and Policy Instrument (NCPI) surveys.
• Sample: A balanced panel dataset covering 8 Southeast Asian countries from 2018 to 2025 (Total observations $N=64$).
• Exclusions: Brunei and Vietnam were excluded because they provided data for only one year ($T_i=1$), which is insufficient for panel estimation.
• Data Processing: Missing data for non-reporting years was handled using Forward Filling (Last Observation Carried Forward - LOCF), assuming policy status remains constant absent legislative change.

Variables:

• Dependent Variable: hiv_prevention_sex_workers (Binary: 1 if an HIV prevention service component targeting sex workers exists; 0 otherwise).
• Independent Variables (Policy Indicators):
  1. barriers_spi: Existence of legal/administrative barriers to social protection.
  2. sp_upaid_care_hiv: Integration of social protection with HIV programs regarding unpaid care work.
  3. sp_agyw: Policies targeting Adolescent Girls and Young Women (AGYW).
  4. sp_coord_nap: Existence of a coordination mechanism/platform including the National AIDS Programme (NAP).
  5. national_sop: Formality of Standard Operating Procedures (SOPs) for sex worker HIV prevention.
• Variable Transformation: Multi-category variables were transformed into binary indicators to measure the impact of high policy levels against a baseline.

Econometric Models:
The study employed Fixed-Effects (FE) and Random-Effects (RE) models to analyze the data.

• Model Selection: A Hausman test (Sargan-Hansen C-statistic) was used to choose between FE and RE. The result ($\chi^2(5) = 235.37, p < 0.001$) rejected the null hypothesis that country-specific effects are uncorrelated with explanatory variables, confirming that the Fixed-Effects (FE) model is the consistent and appropriate estimator.
• Robustness Checks:
  1. Clustered Standard Errors: Applied at the country level to account for autocorrelation and heteroskedasticity.
  2. Lagged Variables: A one-period lagged model ($t-1$) was used to test for reverse causality and determine if policy effects operate with a time delay.

3. Key Results

Descriptive Statistics:

• The most adopted policies were those recognizing AGYW as beneficiaries and establishing coordination mechanisms (Mean $\approx 0.63$).
• The least adopted policy was the existence of national SOPs for sex worker HIV prevention (Mean $\approx 0.16$).
• Multicollinearity was not an issue (Max Spearman correlation = 0.60; Mean VIF = 1.37).

Contemporaneous Analysis (Year $t$):

• Finding: There is a statistically significant positive association between the existence of legal/administrative barriers to social protection (barriers_spi,t) and the presence of HIV prevention services for sex workers.
• Coefficient: $\beta = 0.8531$ ($p < 0.001$).
• Interpretation: In the same year that a country reports barriers to social protection, it is significantly more likely to have an HIV prevention service component for sex workers. This suggests that the recognition of barriers often coincides with the initial prioritization of targeted services.

Lagged Analysis (Year $t-1$):

• Finding: The relationship reverses when using lagged variables. The existence of barriers in the previous year (barriers_spi,t-1) has a statistically significant negative association with current service availability.
• Coefficient: $\beta = -0.3540$ ($p < 0.05$).
• Interpretation: While the initial recognition of barriers may trigger service investment, the sustained presence of these unaddressed barriers acts as a long-term constraint, reducing the availability of HIV prevention services in subsequent years.

Other Variables:

• None of the other policy variables (SOPs, coordination, AGYW focus, unpaid care work) showed statistically significant coefficients in the FE models.

4. Key Contributions

1. First Panel Data Evidence: This is the first study to use fixed-effects panel data to quantify the relationship between national structural policy barriers and HIV service availability specifically for sex workers in Southeast Asia.
2. Temporal Nuance (The "Paradox"): The study uncovers a non-linear temporal dynamic:
   • Short-term: Policy recognition of barriers correlates with service presence (likely due to crisis response and targeted investment).
   • Long-term: Persistent barriers correlate with service decline (structural impediments erode program sustainability).
3. Methodological Rigor: The study addresses endogeneity and reverse causality through lagged variables and clustered standard errors, providing a more robust causal inference than cross-sectional studies.

5. Significance and Implications

• Policy Implications: The findings challenge the assumption that simply identifying or reporting structural barriers is sufficient. While reporting may coincide with initial funding, the study demonstrates that failure to remove these barriers leads to a decline in service availability over time.
• Strategic Shift: National HIV programmes in Southeast Asia must shift from merely adopting targeted programs to the urgent removal of structural barriers (legal and administrative) to ensure long-term sustainability.
• Operational Priority: Barrier removal should be treated as a time-sensitive operational priority integrated into annual reviews, rather than a distant strategic aspiration.
• Global Context: The results support the UNAIDS Global AIDS Strategy (2021–2026), reinforcing that structural barrier removal is a necessary condition for sustained epidemic control.

Conclusion:
The study concludes that while the immediate recognition of structural barriers may trigger HIV service provision, the sustained existence of these barriers is detrimental. To ensure the long-term stability and effectiveness of HIV prevention for sex workers, policymakers must prioritize the active dismantling of legal and administrative impediments to social protection.